aagaard, jacob & shaw, john - experts vs. the sicilian (2004, 1st ed)

Experts vs. the Sicilian

Edited by: Jacob Aagaard

John Shaw

www.qualitychessbooks.com

First published in 2004 by Quality Chess Europe AB Vegagatan 18, SE-413 09 Gothenburg, Sweden

Copyright © 2004

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, electrostatic, magnetic tape, photocoping, recording or otherwise, without prior permission of the publisher.

ISBN 92-9 75243-3- 6

All sales or enquiries should be directed to Quality Chess Europe Vegagatan 18, SE-413 09 Gothenburg, Sweden tel: +46-3 1-244790 fax: +46-3 1-2447 14 e-mail: [email protected] website: www.qualitychessbooks.com

Edited by Jacob Aagaard and John Shaw Typeset: Ari Ziegler Cover Design: Eija Schussler Printed in the Poland by DTL Sp. zo.o.

Content

The writers 6 The Najdorf 9 The Dragon 42 The Sveshnikov 90 The Classical Sicilian 1 1 1 The Kan and Taimanov 148 The Accelerated Dragon 188 The Scheveningen 199 The Kalashnikov 2 19 The Four Knights 228 The Pin Variation 23 6 The Nimzowitsch Variation 248 Minor lines 2 64 yh move alternatives 280 List of variations 28 6 List of games 288

List of symbols

t

!!

!?

?!

?

??

+-

±

;!; =

00

= 00

+

=+=

-+

1-0

YZ-YZ

0-1

(x) corr.

Check A strong move A brilliant move An interesting move A dubious move A mistake A blunder

White has a winning position White is better White is slightly better The position is equal The position is unclear Compensation for the sacrificed material Black is slightly better Black is better Black has a winning position

White won The game was drawn Black won x'th match game Correspondence game

Preface

GM John Nunn had a lot of success with his trilogy Beating the Sicilian 1, 2 and 3. Others have tried to follow suit, but none with the same success. (The latest being Nigel Davies with Taming the Sicilian, where the White repertoire was based on g3-lines against almost everything.)

It was with this knowledge that Ari Ziegler and Jacob Aagaard discussed the idea of a repertoire book against the Sicilian in 2003, while developing the idea of a new chess publishing company. The discussions (leading to this book) end++ed with the idea of contacting strong players who had specialist knowledge in the lines in question. It was our conviction that this would give the reader the best possible insight into the finer points of a particular line. As experienced players and opening book writers we know that important finesses are missed if you do not:

1 . Use a lot of time analysing the games, instead of just believing the players' own analysis. 2. Have prior knowledge of the system. It is obviously not easy to get many busy chess players to deliver up-to-date material

all at the same time. However it was also not as difficult as we feared. Alexander Raetsky and Peter Wells were playing tournaments at the time of the deadline, but still managed to deliver with only a week's delay. And this despite serious computer problems for both!

It has been interesting to learn how differently some very strong players view opening theory, and see how this has made itself apparent in their contributions. At one extreme there is Viktor Gavrikov with his dense theoretical style, at the other Peter Heine Nielsen with his ideas-based approach. This is not a matter of playing strength or necessarily style of play. These two GMs are the two highest rated players contributing to this book, and are both renowned theoreticians. For this reason we decided that it did not make any sense to make huge changes to the style chosen by the different contributors. Clearly a lot of general editing has been done, but we made no particular effort to limit the diversity of the book's authors.

We hope you will find this book enlightening and entertaining.

Glasgow, October 2nd 2004 Jacob Aagaard John Shaw

6 The Writers

The writers

Grandmaster Thomas Luther vs. the Najdorf 35-year-old Thomas Luther from Erfurt in Germany (where Martin Luther went to university) is twice German champion and a regular member of his country's Olympiad team, including 2000 when they were close to winning the tournament, but in the end had to settle for silver medals.

Although this is Thomas' first contribution to a chess book, his 20 years of playing 6.ig5 against the Najdorf at a high level cannot but impress.

Grandmaster Mikhail Golubev vs. the Dragon Mikhail Golubev is a strong 34-year-old grandmaster from Ukraine who mainly considers himself a journalist. He is known as a diligent chess writer and the author of some well-received opening books.

Mikhail contributes often to New In Chess Yearbook with theoretical surveys, and mainly on the Sicilian Dragon. In recent years a great number of books on the Dragon have been published, but none caught the attention of the editors of this book as Golubev's small book, Easy Guide to the Dragon.

We are very happy that Mikhail accepted our invitation to contribute to this book.

International Master Jacob Aagaard vs. the Sveshnikov & several minor lines Jacob Aagaard is 3 1 years old, born in Denmark, but resident in Glasgow, Scotland. His best results are his two GM-norms, both attained in 2004. Jacob has written many chess books. Especially close to heart is the Excelling at Chess series of 5 books, from which the first, Excelling at Chess, won book of the year at chesscafe.com, while the final two received even better reviews.

Jacob is also the author of Easy Guide to the Sveshnikov (Everyman Chess 2000) .

Grandmaster Peter Wells vs. the Oassical Sicilian Peter Wells is 39 years old and has for many years been one of the best players in England. During his work for this book he found time to take second place in the British Championship. At the publication date of this book Peter will represent England at the Olympiad.

Peter's participation is a real scoop for this book. His reputation as a chess opening author is unchallenged. Kasparov, with his usual diplomacy, said about Peter's book on the Semi-Slav, that he could not understand how such a weak player could write such a

Experts vs the Sicilian 7

great book. His recent book on the Trompowsky (Batsford 2003) was called "the finest opening book I've ever seen" by 1M Jeremy Silman, and received universal acclaim as well as a nomination for book of the year at www.chesscafe.com.

In 1 998 Peter wrote The Complete Richter-Rauzer together with Viacheslav Osnos. He plays the Classical Sicilian often and with good results.

Grandmaster Sune Berg Hansen vs. the Taimanov and the Kan Sune Berg Hansen is 33 years old and has been one of Denmark's strongest grandmasters for many years. He has competed in several Olympiads and once in the World Championship. He is well known in Denmark for the high quality of his chess annotations, and as the daily chess columnist for the large newspaper Politiken. His article in this book is his first larger contribution to a chess book. His great knowledge of opening theory will become apparent to anyone who reads his work in this book.

Grandmaster Peter Heine Nielsen vs. the Accelerated Dragon Peter Heine Nielsen is 3 1 years old and currently Scandinavia's number one. Peter has won many international tournaments, ahead of such players as Ivanchuk, Short, Svidler and Beliavsky. He also won a bronze medal at the 1 994 Olympiad in Moscow.

Peter co-authored the book The Sicilian Accelerated Dragon in 1 998 with fellow Dane Carsten Hansen. They are currently contemplating an updated edition.

Grandmaster Vlktor Gavrikov vs . the Scheveningen Viktor Gavrikov is 47 years old and famous on the tournament circuit for his vast knowledge of opening theory. As a player he has competed at the highest level for many years, and won games against players such as Karpov, Beliavsky, van Wely, Lautier, Andersson and Adams. Currently he contributes theoretical articles to ChessBase Magazine and is working on a book on the middlegame.

Viktor has played the Keres Attack with both colours, but does not consider himself a true expert. However his contribution to this book suggests otherwise.

International Master Jan Pinski vs. the Kalashnikov Jan Pinski is a 25-year-old journalist, currently working hard on uncovering corruption in his native Poland, as well as on his next chess book. Jan has written a number of chess books, the first being The Kalashnikov Sicilian with Jacob Aagaard. In an e-mail to the editors Jan states, "It is incredible that I played this line for so long without being punished!"

8 The Writers

International Master Alexander Raetsky vs. the Four Knights. 42-year old Alexander Raetsky very recently made his first grandmaster norm after 9 , 1 0 and 1 1 rounds o f the Biel Open 2004, but was unfortunately given one and not three norms for the effort. He has for a long time been one of the best players in his home region of Voronezh in Russia, where for the last five years he has organized one of the largest open tournaments in the world.

Alexander is also the author of several chess books, most often with his close friend Maxim Chetverik, as well as a contributor to New In Chess Yearbook. Among his books is Meeting 1.e4, which is a repertoire book with the main line being the Four Knights Sicilian, an opening he has played regularly since.

International Master John Shaw vs. several minor lines. John Shaw is a strong 1M from Scotland who has represented his country in many international team tournaments, including Olympiads. He has written two opening books for Everyman Chess, and will soon release a book on the 3 . .tb5 Sicilian.

Experts VS. the Sicilian 9

1he Najdorf

- By 7homas Luther

The Najdorf System is one of the most popular systems of the Sicilian Defence. It arises after the moves 1.e4 c5 2.tLJa d6 3.d4 cxd4 4.tLJxd4 tLJf6 5.tLJc3 a6. The Argentine Grandmaster M. Najdorf played it for the first time in a tournament game in the 40s.

Nowadays it is seen in every level of tournament. Many World Champions, including Fischer and Kasparov, used it as their main defence against l .e2-e4. In many variations an uncompromising battle arises where every move has great importance. It is a very practical choice if Black wants to play for a win from the very beginning of the game. In our times many moves from the older games belong only to history, because strong computer programs show that they are incorrect. Nevertheless, in some lines

White just crushes Black's set-up. I will give some examples where I show the reader some basic ideas (for example the ctJ c3-d5 sacrifice) , and I try to show the connection of different variations and the tricks of move orders.

This book recommends 6.ig5 . I have played this move for nearly 20 years now and I have won many games with it. There are relatively few recent games in the 6.ig5 line, because 6. ie3 is more popular right now. However when comparing the results of these two variations we see that 6.i.g5 is doing fine.

There are some specialists in this line and I have annotated some of their best games. Among many others I want to mention GMs Short, Timman, Kotronias and Sulskis for their great efforts .

The most important lines are the Poisoned Pawn variation (6 . . . e6 7.f4 Wfb6) , which is the most critical line and the mainline (6 . . . e6 7.f4 tLJbd7 8. Wff3 Wfc7 9. 0-0-0 ie7) and now 10 . id3 . These two lines dominate at the moment in tournament practice. Other formerly well known lines, like the Polugayevsky Variation (6 . . . e6 7.f4 b5) , are rarely met nowadays.

I have checked most variations given in this chapter with my computer. But sofr- and hardware are developing fast, and sooner or later improvements will be found. If you are uncertain about

1 0 The Najdorf

a posltlon after reading this book I truly advise you to check it with your computer.

In the beginning I will give some sidelines. Each of them is dangerous if White does not know what to do. I start with 6 . . . �bd7 (the usual move which is played in almost all other games here is 6 . . . e6) . Black's idea is to avoid getting double pawns on the f-line, and maybe later there could be an e7 -e5 in one move. In most of the games Black just plays e7-e6 on the next move and the game transposes to another line. Really not recommendable is this idea in connection with 7 . . . '!Wb6. Black is just too far behind in development to do so. The following game is a perfect example of how White should deal with this plan.

Game l Stripunsky - Granda Zuniga New York 1 998

1 .e4 c5 2.�f3 d6 3.d4 cxd4 4.�xd4 �f6 5.�c3 a6 6 • .tgs �bd7

6 . . . h6 7.ixf6!. 7.£4 '!Wb6?!

7 . . . e6 is of course the move, transposing to 6 . . . e6 7.f4 tLlbd7.

7 . . . h6? 8 .ixf6 tLlxf6 9.e5 dxe5 1 0.fxe5 tLl d5 I l .e6±.

7 . . . b5? also does not really work. After 8 .ixf6 tLlxf6 9.e5 b4 10. tLl cb5! - Nunn. White has a strong position. Here are some lines:

10 . . . tLl e4 I l .e6 '!Wb6 1 2.exf7t md8 l 3 .'!WS d5 14.0-0-0, axb5?! 1 5 .'!Wxe4!+-.

1O .. . dxe5 I l .fxe5 ig4 (Nunn gives 1 1 . . . axb5 12.exf6 gxf6 l3.WlS id7 14.tLle6+-) 1 2J.vd3± axb5 l 3.exf6 exf6 14.\1;Ye4t '!We7 1 5 .ixb5t id7 1 6.ixd7t mxd7 17. tLl e6!!+-. Let's return to 7 . . . '!Wb6?!

8.'!Wd2 '!Wxb2 Otherwise Black's play does not make a

lot of sense. 9.gbl 'lIa3 IO.ixf6!

White uses his lead in development by this immediate action. IO ... gxf6

Forced. 1 0 . . . tLlxf6 I l .e5 tLlg4 1 2. tLl d5 is pretty hopeless for Black. 1 2 . . . \1;Y c5 ( l2 . . . l3a7 1 3J:1b3 '!Wxa2 14.\1;Yc3 id7 1 5 .\1;Yc7+-. 1 2 . . . l3b8 l 3 .tLlc6 \1;Yxa2 14.\1;Ydl ±) l 3 .tLlb3 '!Wc6 14. tLl a5 \1;Yc5 Now the weaker player could have won if he played 1 5 .tLlxb7+-, but respect for the grandmaster made him repeat moves, Bindrich - Zagrebelny, Dresden 2000.

Experts VS. the Sicilian 1 1

1 1.�d5 �b8 Other moves are no better, or maybe

even worse. 1 1 . . . �xa2 12 . �b4 @dS ( 1 2 . . . b5

1 3. �c7t @dS was Zunker - Holfelder, Bruchkoebel 2002. Now 14.�xaS ib7 1 5 .i.c4! wins) 1 3 J �b3 �al t 14. @£2 e6 1 5 .E:b l �a2 1 6.i.c4 a5 17.�c3 ltlc5 I S. ltl c6t bxc6 1 9 .�:xf6t @eS 20. ltlc7t @d7 2 1 . �xf7t i.e7 22.ixa2 1-0 Garbarino - Sabas, Buenos Aires 1 9S2.

1 1 . . .�c5 1 2. ltl b3 �c6 1 3 .ltla5 �c5 14. ltlxb7 E:bS 1 5 .ltlxc5 E:xb l t 1 6. @£2 ltlxc5 17.�a5+-. 12.�b3!

White needs to bring his pieces into action. Worse was 1 2. � c7t @dS 13 .ltlxa6 bxa6 1 4. ltl c6t @c7 1 5 . ltlxbS ltlxbS 1 6.ic4 ltl c6+ .

12 ... Yla4 After this there is not a lot to talk about.

White is simply much better. 1 2 . . . �c5 13 . E:c3 �a7 14 .ixa6!+-

does not work, but 12 . . . �xa2!? has been suggested, and is in fact the only way for Black to play on. Still, analysis assisted by a computer indicates that White has the advantage. 1 3 .ic4! is of course the move. Now we have:

a) 1 3 . . . �al t? This only helps White. 14 . @£2 �xh 1 . White now has a winning combination with 1 5 . ltl c7t @dS 1 6.�a5! b6 17. ltl de6t fxe6 I S .ltlxe6t @eS 19 .�h5 mate.

b) 13 . . . ltl c5? 14Jhb7! ( 14.0-0! ? ltlxb3 1 5 .ixb3 �a3 1 6. E:f3, is also tempting, but winning the queen is more convincing.) 1 4 . . . �bl t 1 5 . E:xb l E:xb l t 16 .@e2 E:xh1 17.�a5+-.

c) 13 .. . e6 14 . ltl c7t @dS 1 5 .�c3! ( 1 5 .0-0 �a4! and it is not possible to find more than equality for White. This shows the old truth that an advanced soldier behind enemy lines can do a lot of damage.) 1 5 . . . ltl c5 1 6. E:xb7 �blt 17 . E:xb l E:xb l t I S . @e2 E:xhl 1 9 .�a5 @d7 20. ltl cb5! and the White attack crashes through.

d) 1 3 . . . �a4 14 .�c3 ltlc5 1 5. ltl b6 �a2 1 6.0-O ltlxe4 17.�el ltlc5 lS .�b4 and the black queen is trapped. 13..ha6! e5!?

1 3 . . . bxa6 14 . � c3!+- Vitolinsh - Arakas, USSR 1 975.

13 . . . �xa2 14 .�c3 e6 1 5 . ltl c7t @dS 1 6.ic4 is of course not playable for Black. There is nothing that justifies the weakening of the king's position. 14.gb4 Ylxa2

1 2 The Najdorf

14 .. .'�a5 1 5 .ttJb3 looks good for White. 1 5 . . . 1M/xa2 1 6 . .ib5 with a crushing attack. 15.�b3 bxa6? 16.ti'c3! 1-0

Black resigned. His queen is trapped after 1 6 .. J'1a8 17.0-0 a5 1 8 . Eib5 a4 1 9 .ttJc7t 'it>d8 20. ttJcl .

Another old sideline is 7 . . . �c6. It came to popularity after GM Shabalov played it. Funnily it was also GM Shabalov who started crushing this line.

Black wants to achieve a Rauzer-like setup and make use of White's early £2-f4. In fact the early f4 gives White the chance to kick Black's knight on f6 with e4-e5 . Since White has better development the tactics should go fine for him, and they do so. In the game below GM Adams shows fine technique and gains a great advantage. Only a silly blunder, which had nothing to do with the opening, cost him half a point.

Game 2 Adams - Anand Linares 1 997

1 .e4 c5 2.�f3 d6 3.d4 cxd4 4.�xd4 �f6 5.�c3 a6 6.i.g5 e6 7.f4 �c6

Usually in the Najdorf the ttJ b8 is going to d7, compared to the Rauzer where Black sets up with, d6, ttJ f6 and ttJ c6. 8.e5!

Here Black wants to make use of the

early £2-f4, so after the "normal" Rauzer move 8 .'�d2 h6 9 . .ih4 (9 . .ixf6 ti'xf6 is not attractive for White either) 9 . . . ttJxe4 is very strong. But this is not a Rauzer, but a Najdorf, where White is prepared for an early e4-e5 . 8 ... h6 9.ih4

9 . • • dxe5 9 . . . ttJxd4 keeps the pawn structure on

the queenside intact, but Black has to commit his kingside with g7 -g5 . 1 0.1M/xd4 ( 10.exf6?? falls into a nice trap: 1 0 . . . ttJf5! This was discovered by Adorjan. I I .fxg7 ti'xh4t 1 2.g3 ttJxg3 1 3 .gxh81M/ ttJ e4t and notwithstanding his two queens, the white king will soon be checkmated.) 1 0 . . . dxe5 1 1 .1M/xd8t 'it>xd8 1 2.fxe5 g5 1 3 . .ig3! ttJ d7 14.0-0-0 .ig7 1 5 . .ie2 (for some reason this natural move is not in Kosten's book Easy Guide to the Najdor/J 1 5 . . . 'it>e7 16 . .ih5! (f7 is the weakest point in Black's territory) 16 . . . EifS ( 1 6 . . . ttJxe5 does not work here. 17 . Eihel f6 1 8 . .ixe5

Experts vs. the Sicilian 1 3

fxe5 1 9 . 8:f1 ! !f6 20.it:)e4 8:f8 2 1 . 8:f2! and White is clearly better.) 17.it:)e4 it:)xe5 1 8 . 8:he1 f5 1 9 .it:)c3 f4 20.!f2 b6 2 1 .!xb6 !b7 22.!c5t 'it>f6 23. 8:xe5 1-0 Luther - Senff, Cappelle la Grande 200 1 . 10.�xc6 'i!Yxdl t l l .:axdl hxc6 12.fxe5 �d5

Also possible is 1 2 . . . it:)d7!? but Black still has a passive position. 1 3 .it:)e4 g5 14.!g3 �g7 1 5 .it:)d6t 'it>e7 1 6.it:)c4 a5 17.h4. This is a very strong move: White wants to weaken the g5-pawn and trade his passive rook. Now it is difficult for Black to develop his last pieces, as can be seen by the following lines: 17 . . . a4 07 . . . !a6 1 8 .it:)xa5 8:hc8 19 .!xa6 8:xa6 20.it:)c4 8:xa2 2 1 .hxg5 hxg5 22.8:h5 8:h8 23. 8:xg5 ±) 1 8 .hxg5 hxg5 19 . 8:xh8 !xh8 20.!e2;!; . 13.�e4 B:h8 14.h3

14.c4?! allows a tricky piece sacrifice 14 . . . 8:xb2! 1 5 .cxd5 !b4t 16.it:)d2 exdH.

14 ... !e7 14 . . . g5?L Making the check on b4

happen is not worth weakening the structure. 1 5 .�g3 !b4t 16 .'it>e2! and by threatening c2-c4 White obtained a big advantage in Brodsky - Rechel, Groningen 1 993. 15.1g3!

This move certainly secures an advantage for White. Black has too little space for his pieces. This is more important than just the usual good/bad bishop stuff. Worse is 1 5 .!xe7 'it>xe7, and with a weak pawn on e5, White can never be better. 15 ... 0-0 16.i.e2

1 6.c4?L White should be careful with pawn moves: 1 6 . . . it:) b4 17 . 8:d2 8:d8 with counterplay. 16 ... a5

1 6 . . . it:)e3 hunting the g2- pawn is not good for Black: 17. 8:d2 it:)xg2t 1 8 . 'it>f2 it:)h4 1 9.it:)f6tL Without this move White would have nothing. 1 9 . . . gxf6 20.!xh4 !c5t 2 1 . 'it> 8 fxe5 22.!f6 and after �d3 and 8:g2 White has a dangerous attack. 17.c4 �b4 18.EM2!

It is important to protect the a-pawn since it keeps Black's knight out of the game. Huzman gives: 1 8 .0-0 it:)xa2 19 . 8:a1 it:)b4 20. 8:xa5 it:)c2 and Black has good counterplay against the b3-pawn. 18 ... B:d8 19JUl!

Another brilliant move by Adams. The king stays in the centre to cover the important squares. 19 .. Jixd2 20.<it>xd2 �a6

14 The Najdorf

20 . . . tlha2 Now this is different. The white king dominates the knight on b4 after: 2 1 .a:al lLlb4 22.a:xaS the position is ± . 21 ..th5

Forcing Black to weaken the kingside structure. 2l . .. g6 22..if3 .tb7 23.i>c3 gd8 24.�d6 .taS 25.a3??

This spoils all the previous achievements. After protecting the knight on d6 once more, White's victory would have been only a question of time. 2S.a:d l ! was the right move.

25 ... £5! Now Black has counterplay.

26.b4 g5 27.h3 .tm 28.c5 gb8 28 . . . ig7 29. lLlc4 does not change

much. 29..th5

29.a:al with the idea 30. lLlc4 was recommended after the game. 29 ..• �c7 30..ta �a6 31 ..th5 �c7

32.1a lh-lh White could have played on, but

probably he was frustrated with his 2Sth move.

Now we turn to the above mentioned game from GM Shabalov. Instead of heading for an ending, as in the previous game, Black can burn his bridges and crack White's centre with 9 . . . gS. White has to play carefully and have some theoretical knowledge about the position. With the right move order White can prevent Black building up a strong centre. He has to take the dS knight before he takes the one on c6. As soon as White castles the black king will find himself under a strong attack.

Game 3 Shabalov - Browne Las Vegas 1 997

l.e4 c5 2.�a d6 3.d4 cx:d4 4.�xd4 �f6 5.�c3 a6 6..tg5 e6 7.£4 �c6 8.e5! h6 9..th4 g5

Black acts concretely against White's centre and the game becomes very tactical. Since White has better development, tactics should work in his favour. lO.fxg5 �d5

1 O . . . lLlh7 Knights need to be placed in the centre! l 1 . lLl xc6 bxc6 1 2.exd6 with advantage for White.

Of course not 1 0 . . . hxgS?? l 1 . lLl xc6 bxc6 12 .ixgS and Black loses a piece.


1 1 .c!Llxd5 It is important to take first on d5 and

later on c6, so that Black does not have the option to take with a later c-pawn on d5. n ... exd5 12.exd6

The best. Other moves like e5-e6 have been tried, but without much success.

12 ... Yfxd6 12 . . . !xd6 1 3 .c!Llxc6 bxc6 14 .�d4

White plays this move because he wants to protect his bishop on h4 and then play g5-g6. 14 . . . �e7t 1 5 .!e2 !e5 1 6.'1INa4 Elb8 ( 1 6 . . . �d6 is met by a surprisingly strong move: 17.ig3 !± id7 ( 17 . . . ixg3t 1 8 .hxg3 Wfxg3t 19 . Wfl Both c6 and h6 are hanging, besides Black's king is in danger. White is dearly better.) 1 8.gxh6 with advantage for White.) 1 7.g6 (when White achieves this Black is usually busted) 17 . . . �d6 ( 17 . . . �b4t. This was once recommended as equalising, but I think White is still better. 1 8 . Wfxb4 Elxb4

1 9.9xf7t Wxf7 20.0-0t Wg7 2 1 .!f2 Elxb2 22.!d3. After lots of exchanges Black still has problems, his pieces are not coordinated. A possible line could be: 22 . . . ElfS 23.Elael !c3 24.Ele7t Wg8 25 .!c5 l3xfl t 26. Wxfl !g4 27.Elc7 Elxa2 28.Elxc6 !dI 29.!e3 Ela3 30.Elxa6 Elxa6 3 1 .!xa6 !xc2 32.!xh6 and good technique should bring White the full point.) 1 8 .gxf7t WfS 1 9.!f2 Elxb2 20.0-0 Elh7 2 1 . Wh l Elxf7 22.id4 With a decisive attack in McDonald - Danner, Budapest 1 996. 13.Yfe2t

The point of White's play. White gets the dearly better game now. 13 . . .i.e7 14.c!Llxc6

Of course not 14.0-0-O?? �f4t-+ . 14 ... bxc6 15.i.g3! Yfg6

1 5 . . . �b4t 1 6.c3 and, thanks to Wfe2, the b2 pawn is protected. 16.Yfe5! l3g8 17.gxh6!

1 6 The Najdorf

Finally! White secures his extra pawn. 17 .. . YlYxc2

1 7 . . . !f5!? I B .!e2! !xc2 1 9.0-0 and with his king in the centre Black is helpless against all the threats.

1 7 . . . �xh6 (with this move Black is just accepting to play a pawn down) I B .!f4 �g6 ( l B . . . �e6 1 9. �xe6 !xe6 20.g3 was seen in Luther -Abreu, Havana 200 1 . Black had no compensation for the pawn.) 1 9 .0-0-0±. With a safe king and an extra pawn White has a dear advantage. 18.ie2 �gS

Desperation! I B . . . �g6 19 .!h4! is very uncomfortable for Black. 19.YlYh8t! 'it>d7 20.YlYc3+-

This finishes all Black's hopes. 20 ••• YlYxc3t

20 . . . �e4 2 l .0-0 �xe2 22. l':iae l and the various threats cannot be parried anymore. 21.bxc3 i.f6 22.0--0

After this move everything is dear. The passed pawn on h6 decides the game. 22 ..• ,ixc3 23.�xf7t 'it>e6 24.�af1 id7 2S.i.h4 gg6 26.ihS 1-0

Now after l':ixh6 !g4t White wins a piece, so Black resigned.

The early �c7 is another sideline. Black wants to play b7-b5 without allowing e4-e5 . If White does not react to this plan and slowly develops, Black will kick White's knight on c3 by playing b7-b5-b4. If White has to remove this knight to e2 or a4 then Black is doing

fine. There are many tactical lines but I cannot recommend them. Basically, if White gets the chance to rake on f6 and Black has to recapture with the g-pawn White should do it. The arising position is more common in the Rauzer Defence, so I advise the reader to study this chapter as well.

Game 4 Khalifman - Lautier Moscow 200 1

1 .e4 cS 2.otS d6 3.d4 cxd4 4.otxd4 otf6 S.otc3 a6 6.igS e6 7.£4 YlYc7

With this move Black wants to trick White in his set-up. If he goes for �dlf3 and castling queenside Black quickly plays b7-b5-b4. Since at this early stage of the game there is no ttJ d5 - sac possible the c3-knight has to be moved backwards, which is a big concession.

8.i.xf6

Experts vs. the Sicilian 17

On the other hand there is the chance to break Black's pawn chain, since Black has not played either �bB-d7 or �f8-e7. In my opinion, this is the most principled way to treat the 'W c7 -line.

B :�f3 is often played in this position. B . . . b5 (Black decides not to enter one of the main lines by playing B • . • ttl bd7 or �e7.) 9 . . bf6 gxf6 1 0 .e5 d5 ( 10 . . . �b7 1 1 . Wfh5 with the idea of ttl d4xe6 and White is better here.) I I .exf6 b4 12 . ttlxd5 exd5 1 3.0-0-0 �b7 and, after studying this position for some time, I came to the conclusion that White should not risk this piece sacrifice. 8 ... gxf6 9.tyd2

9.�e2 is another way of setting up the pieces for White. Generally I do not think the white king belongs on the kingside. 9 . . . ttl c6 1O . ttl b3 b5 1 1 .0-0 �b7 and Black will castle queenside and aim for the standard break d6-d5 . 9 ... b5

Pushing the b-pawn is in the spirit of the variation. The drawback is that the black king will never find a safe spot on the queenside.

9 . . . ttl c6 1 0.0-0-0 id7 1 1 .'it>b l h5 12 .�c4 0-0-0 1 3 .ttlxc6 'Wxc6 14 .�b3 'it>bB 1 5 J �hf1 ;!; was seen in Topalov -Anand, Dortmund 1 997. 10.1d3

1 0.a3 �b7 1 1 .�e2 with the idea of castling kingside is another option, but Black can even stop this plan by playing 'Wc7-b6.

10 ... 1b7 1 1 .0-0--0 �d7 Black cleverly keeps the knight because

it will be strongly placed on c5 . After 1 1 . . . ttl c6 12 . ttlxc6 White is better. 12.:B:hel 0-0--0 13.£5 �c5 14.a3

White has to secure the c3-square for his knight. 14 ... g;,b8 15.g;,bl h5 16.tye3 ih6 17.tyh3 tye7

18.tya ¥2-¥2 Here the players agreed a draw. In my

opinion White could have continued the game. Instead of l B . Wff3 I prefer:

a) I B .b4. OK, it is not everybody's taste to open one's king, but getting rid of the c5 knight is worth it. I B . . . ttlxd3 19 Jhd3 :B:deB 20J l:edl with pressure. Or:

b) I B .�e2, hitting h5. I B . . . e5 ( 1 B . . . ttl xe4? 1 9 .ttlxe4 �xe4 20.if3 and White wins) 1 9. ttl b3 ttlxe4 20. ttl d5 �xd5 2 1 . :B:xd5 and White has good compensation. He has play on the light squares and against Black's king.

1 8 The Najdorf

Great players have their own openings is an old saying which is difficult to fulfil in our time, since most of the sensible (and even most of the stupid) moves from the starting position have been played already. However GM Polugayevsky invented 7 . . . b5, played it and published a lot of analysis on it, so this system took his name: the Polugayevsky Variation. The idea is to kick the c3 knight as early as possible. White is forced to take counter measures and the game develops in a very tactical way. What was a tremendous workload back in the 70s and 80s can now be done quickly with a computer program. Nevertheless the work of GM Polugayevsky will always be remembered.

Game 5 Leko - Ghaem Maghami Yerevan 200 1

1 .e4 e5 2.�f3 d6 3.d4 cxd4 4.�xd4 �f6 5.�c3 a6 6.!g5 e6 7.£4 h5

This aggressive move was played and analysed deeply by the Russian GM Lev Polugayevsky. 8.e5

The only way to deal with b7-b5 successfully. Otherwise Black just manages to kick White's knight with b5-b4. 8 ... dxe5

8 . . . h6? (this move does not promise Black much here) 9 .!h4 g5 10.fxg5 lD h7 1 1 . 'Wh5 hxg5 12.Ag3 Ag7 1 3 .0-0-0 and White had a big advantage in Kasparov -

Ehlvest, Baku 1 978. 9.fxe5 'fIe7

The idea behind Black's play. He does not lose material, but White gains a lot of time. 10.exf6

There is also 10 . Vff e2 but it is not in the spirit of White's set-up. 10 .•• 'fIe5t 1 l .!e2 'fIxg5 12.0-0

12 .'Wd3 is considered as the main alternative here. Bringing the white king out of the line of fire is, in my opinion, the better option.

12 ... ga7 Black's defence is based on this idea:

the rook goes to d7. 12 . . . Vff e5 was for a long time considered

the main line, but is now less popular in practice. It is considered in the next game.

Not the natural 1 2 . . . !b7? 13 .Af3 when Black's position cannot be saved:

a) 1 3 . . . Axf3 14. Vffxf3 ga7 ( 14 . . . Ac5


1 5 .'�'xa8 as in Bisset - Martinez, e-mail 1 994 offers Black no compensation) 1 5 . E:adl 'i;!fe5 16 .E:fe 1 'i;!fxf6 17.'i;!fg3 lLld7 1 8 . lLl d5 'i;!fd8 1 9 .1Llc6 and White wins.

b) 1 3 . . . 'i;!fe3+ 14 .'it>hl �xf3 1 5 . E:xf3 'i;!fe5 16 .�d2! �d6 1 7.g4 b4 1 8 .lLlf5!! After this Black is done for. 18 . . . ic7 19 .E:e 1 'i;!fxh2+ 20. 'i;!fxh2 ixh2 2 1 .fxg7 E:g8 22. lLl d5 lLld7 23. 'it>xh2 'it>d8 24. lLlde7 1-0, Stripunsky - Jaracz, Poland 1 995 .

c) 1 3 . . . E:a7 14 . lLlxe6!! An absolute stunner. 1 4 . . . fxe6 ( 14 . . . �e3+ 1 5 . 'it>hl fxe6 1 6.fxg7! and i t i s the end of the world) 1 5 .£1+ 'it>e7 1 6.'i;!fd4 The key move. At best Black will be an exchange down with a ruined position. 16 . . . 'it>xf7 17.ixb7+ 'it>e8 1 8 . lLl e4 E:xb7 19 .1Llxg5 E:d7 20. 'i;!f e5 1-0, Kaehmann - Hamburg, Ruhrgebiet 1 999. 13.YlYd3 gd7 14.lLle4 YlYe5

14 . . . 'i;!fd5? ! is worse than the text. The game Sulskis - Stocek, Isle of Man 2002 went as follows: 1 5 .c3 lLlc6 1 6. lLlxc6 'i;!fxc6 17.'i;!fe3 �b7 1 8 .if3 and Black still could not free his position from White's attack. 15.lLlf3!

1 5 .c3?! �b7 1 6.if3 ixe4 17.ixe4 gxf6 and White does not have enough compensation for the pawn. 15 ... YlYxb2

As in many tactical lines Black is forced to take some material.

After 1 5 . . . 'i;!fc7 1 6.�e3 ib7 17.c4

ixe4 1 8 .�xe4 gxf6 1 9 .cxb5 'i;!fb6t 20. 'it>hl axb5 2 1 .a4! . White simply has a great attack for no risk at all. The game Vasquez - Arancibia, Maipu 2003 was soon 1-0 . 16.YlYe3 .tb7 17.a4 b4

After 17 . . . �b4 Black could not solve all his problems following 1 8 .c4 ixe4 19 .�xe4 �c5t 20. 'it>hl b4 2 1 .'i;!ff4 id6 (a serious commitment, but 2 1 . . .lLl c6 22. E:adl 'i;!ff5 23.�e3 E:xdl 24. E:xdl 'i;!fxf6 25 . �b6 is no fun either) 22.fxg7 E:g8 23. �h6 ie5 24. 'i;!fxh7 E:xg7 25 .'i;!fh8t and soon 1-0 in Rodriguez Cespedes - Stangl, Biel 1 988 .

17 . . . �xe4 does not solve Black's problems. 1 8 .�xe4 ic5t 19 . 'it>hl gxf6 20.axb5 and White wins back the material and keeps a clear advantage. IS.gabl

1 8 .c3 This move i s an old recommendation. The text is better. 1 8 . . . ixe4 1 9 .�xe4 gxf6 and the best

20 The Najdorf

White can get is a repetition by following Black's queen with his rooks. 18 .. .'�xc2

1 8 . . . 'IMi'a3 1 9 .c3 Only now does White play this move. Black cannot finish his development and is in trouble. 19.�fg5! V!!c7

Black is in serious trouble as any computer shows. Nowadays any program can analyse this tactical position far better than any human.

19 . . . h6!? does not help either. 20. Eibel 'lMi'xel 2 1 . Eixel hxg5 22.'lWg3 It'ic6 23 . .txa6 and White wins.

19 . . . g6 20. Eifel 'lWa2 2 1 ..tc4 and White wins again . . .

20.gxb4! Opening up the posmon, after this

blow there is no longer a defence. 20 .. .he4 21 .�xe4 hb4

Allowing a nice finish. 22.fxg7 Eig8 23.�f6t d8 24.�xg8 .ic5 25.�f6! he3t 26.c;t>hl c8

27.�xd7 1-0

In the next game we shall continue analysing the Polugayevsky Variation. Compared to a line like 7 . . . Nbd7 it might seem less relevant. But first of all many club players really like to play this way, as there is something macho about it. Secondly, the knowledge necessary for playing an opening is not necessarily always centered around the critical lines.

Game 6 Wosch - Nordin e-mail 200 1

Sometimes a relatively weak player (here 2000 elo) plays at the level of a grandmaster for the entire length of a game. This is the case with this wonderful game. Some might think that this is because of computer assistance, as it is an e-mail game, but looking this game over with my own computer does not suggest this at all. On the contrary! 1 .e4 c5 2.�a d6 3.d4 cxd4 4.�xd4 �f6 5.�c3 a6 6 . .igS e6 7.f4 b5 8.e5 dxe5 9.fxe5 V!!c7 10.exf6 V!!e5t n.J.e2 V!!xg5 12.0-0 V!!e5 13.�a i.c5t

1 3 . . . 'lWxf6 14.lt'ie4 'lMi'xb2 1 5 .lt'ifg5 and, with most of his pieces in the starting position, Black is helpless against White's various threats.

1 3 . . . 'lWe3t 14 . �hl It'id7 1 5 .Eie 1 'lWa7 16.fxg7 .txg7 17.'lWd6 was played in a blindfold rapid game between Leko and


Ivanchuk. I t seems that White still has some pressure here. 14.�hl �xf6 15.�e4 �e7 16.�e5!

The most dangerous. 1 6. ttl fg5 f5! has proven to be nothing. 17 .ih5t g6 I S .ttlxh7 @m and Black was OK in several games.

16 ... f5? Now this does not work. 16 ... 0-0 is

the only move. Now White should play 17. ttlxf7! and then we have:

a} 17 . . . ttl c6? I s .ih5 !± is no good for Black. I have analysed the following line I s . . . id4 1 9 .c3 ie5 20.�g4! with a winning attack. 20 . . . id7 2 1 . ttlh6t @hS 22.l"H7 l:hf7 23. ttlxf7t @gS 24. ttl h6t @hS 25. ttl g5 g6 26Jl:fl ! !l:f8 27. ttl hf7t @gS 2S. �h4 and Black has no defence.

b} 17 . . . ib 7 is an interesting alternative. I s .id3. Nunn's suggestion. ( 1 S .ttlxc5 'lWxc5 19 .'lWd6 'lWxd6 20. ttlxd6 !l:xfl t 2 1 . !l:xfl id5= was played in Bartoli -Innorta, e-mail I 99S) I S . . . !l:xf7 19 .!l:xf7

'lWxf7 ( 1 9 . . . @xf7 20.'lWh5t±) 20. ttlxc5 id5 2 1 . ttl e4 'lWg6 22.'lWe2 and I think White has good chances for achieving an advantage here. He has ideas of ttl e4-c3 and a2-a4, creating further weaknesses in the Black camp.

c} 17 . . . !l:xf7 I S . ttlxf7 @xf7 19 .ih5t @gS ( l 9 . . . g6 20. ttlxc5 !l:a7 2 1 . ttl e4 @g7 22.if3 !l:d7 23.�ea Kover - De Almeida, corr. 1 9S0.) 20. ttlxc5 Now we have the following options:

c l } 20 . . . �xc5 2 1 .�dSt 1-0. Lukas - Feist, corr. 1 997. Black probably overlooked 2 1 . . .�f8 22.if7t!+-

c2} 20 .. . ttl c6 2 1 .�f3 id7 22. ttlxd7 �xd7 23. !l:dl ± Schneider - Riedmueller, corr. 1 996.

c3} 20 .. . ttl d7? 2 1 . ttlxe6!± Beliavsky -Polugaevsky, Moscow 1 979.

c4} 20 .. . !l:a7 2 1 . ttl d3! . A new idea, but not a very surprising one, as the alternatives are less encouraging. (2 1 . ttl e4 !l:d7 22.�e2 ttl c6 23.c3 ttle5 24. !l:fl ib7= Denaro - Bosco, corr. 1 990, and 2 1 .�d4 !l:c7! 22. ttl e4 !l:xc2 23. !l:fl ttl d7! + Mauro - Soranzo, corr. 1 990.) 2 1 . . . ttl c6 22.if3 ttl d4 (22 . . . ib7? 23 .'lWgl ! !l:aS 24. ttl c5±) And now White has many ways to proceed. 23.ie4;!; is probably easiest. Of course Black can fight for a draw in such an endgame, he is only slightly worse, but certainly White would accept this position from the opening. 17 . .ih5t g6 18.�xg6 hxg6 19 . .bg6t �f8 20.�xc5 gh6!

The alternatives are not cheerful.

22 The Najdorf

20 . . . ..t>g7 is met strongly with 21.ttlxe6+! .!xe6 (No better fate is to be found after 21. . . ..t>xg6 22.ttlf4t! (22Jhf5?! ..t>xf5 23.'?;V d3t 'itixe6 24J '!el t ..t>f7 25 .l: he7t 'itixe7 26.�e4t ie6 27.�xaB;j;) 22 . . . 'itig7 (22 . . . 'itih6 23.�h5t+-) 23 J'!f3! . This manouevre is not that easy

to find, but very logical. Black has no way to bring his pieces to the defence of the king. 23 . . . �e5 (23 . . . lMfd7 24. 1"!d3 �c6 25 . 1"!c3 IMfd7 26.1Mff3+-) 24. 1"!g3t 'itif6 25 .�f3 1"!h6 26. ttl h5t 'itie6 (26 . . J !xh5 27. �xh5 and Black has no way to survive the attack.) 27. 1"!dl and the black king cannot escape.) 22.ixf5 IMfh4 (22 . . . ixf5 23. 1"!xf5 1Mfh4 24.lMfd6! transposes) 23.�d6 ixf5 24. 1"!xf5 1"!eB 25 .1"!afl1"!a7 26. 1"!5f4 �dB 27.�c5 1"!c7 2B.�h5 1-0. Uboldi - Lalanne, San Antonio de Padua 200 1 .

2 0 . . . �xc5 2 1 .lMfdBt 'itig7 22.�g5 with a winning attack. A crucial line is 22 . . . 1"!xh2t 23.'itixh2 lMfe5t 24.'itigl �f6 25.�g3 �xg6 26.�c3t+-. 21 .1h5

Probably the best move. 2 1 .lMfg4!? IMfxc5 22.lMfg5 1"!xg6

23.�xg6 1"!a7°o. 2 1 . ttlxe6t ixe6 22.ixf5 if7 looks

unclear to me. 21 ••• �c6!

2L..lMfxc5? 22.�dBt 'itig7 23.1"!f3 1"!xh5 24. 1"!g3t 'itif7 25.lMfgBt 'itie7 26. 1"!g7t 'itif6 (26 . . . 'itid6 27.�f8t 'itid5 2B .1"!dl t and White wins the queen.) 27.1"!f7t 'itie5 2B.1Mf g3t 'itid5 29.l'!dl t

and it is all over.

22.M! This move makes a lot of sense: White

sacrifices his extra pawn to derail the knight. If this or the alternative 22. ttl e4!? is stronger I do not know. The position needs a lot of independent analysis before anything can be said with certainty. I have tried to give some variations here that I believe are critical, however they cannot be said to be conclusive in any way.

22 . . . ib7 (22 . . . e5? ! is the computer's first choice, but after 23.g4!? Seems strange, but it works! 23 . . . til d4 24.lMfd2 IMfg7 25.c3 ib7 26. 1"!ae l ! White has a very strong attack) 23.�g4!? tild4 (23 . . . 1"!dB 24. tilg3 til e5 25 .�f4 �g7 26. 1"!ae l 1"!xh5 27. 1"!xe5 1"!h6 2B. 1"!fel ±) This position is probably critical. I have tried to oudine the possibilities here, but cannot give full conclusions.

a) 24. 1"!adl !? tilxc2 25 . 1"!d3 1"!h7 {25 . . . id5 26. 1"!g3 �h7 27. tilg5 IMfd7


28. tLlh3! Wih7 29.�g6 gxg6 30.�xg6 �xg6 3 1 . gxg6;!;) 26. tLlg5 @g8 27.�h4 gg7 28. gg3 gf8=.

b) 24.c3? tLlc2 25 . gadl tLl e3 26.Wif4 gxh5 27.Wixe3 Wih4-+.

c) 24. tLlg3! �f6 25 .Wif4 e5 26.Wif2;!; The following analysis might be correct, but chances are that they are a bit too long to be bulletproof. 26 .. .f4 27.c3 tLl e6 28 .gadl gd8 29.Wib6 gxdl 30.�xdl �xg2t. Far from the only option here. 3 1 . @xg2 gxh2t 32. @gl ! ? (32. @xh2 �h4t=) 32 . . . gd2 33.�b3 Wih4 34. gf3 �h3 35 . �b8t @e7 36. Wib7t @f8 37. tLl e2 Wig4t 38 . @f2 �h4t 39. @f1 gxb2 40. gd3± . Of course all of this is not that clear, and improvements can probably be found for both players. 22 .. .ti::Jxb4?

After this I cannot find a good position anywhere for Black. 22 . . . a5! looks stronger. 23 .�f3 Wic7 24.h3 axb4 25 .Wid2 gg6 (25 . . . @g7 26.c3 looks dangerous) 26. tLl d3;!; . 23.ia Vlfc7 24.h3 �c6 25.a4! bxa4

Also after 25 . . . b4 26. Wi d2 Black has no easy choice. 26 . . . gg6 (26 . . . @g7 27.c3±) 27. tLl d3 �b7 28. tLlxb4± . 26.Vlfd2 �g7

It is close to being over for Black. The following line clearly shows the potential in White's position. 26 . . . gg6 27.�h5 gg7 28.Wih6 �e7 29. gae l ! @g8 30. tLlxe6 �xe6 3 1 . gxe6+-. 27J�xa4 e5

27 . . . ga7 28. gh4! loses for Black

as well, but it might have taken a few extra moves. 28 . . . gxh4 29.�g5t @f7 30.�xh4 and the black king i s simply too fragile. One line could be 30 . . . tLl e5 3 1 . Wih7t @f8 32.Wih6t @e7 33.�g5t @f8 34.Wif6t tLlf7 35 .tLlxe6t .txe6 36.Wixe6 Wixc2 37.�d5+-. 28J�h4 gg6

28 . . . gxh4 29.�g5t @f8 30.�f6t! � f7 3 1 . � xh4 and everything is going wrong for Black. 29.J.h5 1-0

Black resigned. Probably a bit early, but after 29 . . . Wid6 30. tLld3 �f6 3 1 . �xg6 Wixh4 32.�xf5 White has a very strong attack and an extra pawn.

For a long time the set-up with I O.g4 was considered as the main variation against the main line with 7 . . . �e7. There is still nothing wrong with it and maybe it will once again become the main line in the future. However, right now I O.�d3 enjoys greater popularity. This will be discussed in another game. Most of the games in the I O.g4 line are very old. When 1 3 .5 was discovered for many years the line saw a revival and much analysis was published. Seeing it from today's point of view I have to say that the line still has lots of resources for both sides.

24 The Najdorf

Game 7 Luther - Efimenko Ohrid 200 1

In this game I give a summary of older lines, which have gone out of fashion. Like anything old, there may one day be a revival. 1 .e4 c5 2.�a d6 3.d4 cxd4 4.�xd4 �f6 5.�c3 a6 6 . .ig5 e6 7.£4 .ie7

7 . . . id7 is an interesting concept. Black wants to put his knight on c6, after which the position has a more Rauzer-like character. 8. � e2. White takes the chance to place the queen on e2. In the Rauzer the queen is less efficient on d2. 8 ... � c6 9.0-0-0 lLlxd4 (9 . . . �c7 1 0.ixf6 gxf6 1 1 . \tl b 1 gives White a typical position where he is slightly better) 1 0.Eixd4 Wfa5 ( l 0 . . . Wfb6 This does not give Black equality either. 1 1 . � d2 ic6 This was played in Najditsch - Nakamura, Wijk aan Zee 2004. Now 12 .ixf6 gxf6 13 .ic4!;!; and a future f4-5 will cause Black a lot of problems.) 1 1 . � d2 ic6 1 2.5 e5 13 .ixf6 exd4 14 .ixd4 b5 1 5 .�e3 with a big advantage for White in Lastin - Cvitan, Moscow 200 1 . 8.�a V!!c7

8 . . . Wfa5?! This line has recently been refuted by Radjabov. 9 .0-0-0 id7 10 .ixf6! (the older move 1 0.e5 does not give White anything) 1O . . . ixf6 l 1 .e5!

Now the threat of �xb7 causes Black trouble. 1 1 . . .dxe5 1 2.fxe5 ig5t 13 . \tlbl lLlc6 14. lLl e4 lLlxe5 ( l 4 . . . �xe5? ! . This only opens up Black's king. 1 5 . lLl 5 exf5 16 .lLlxg5 �e7 17.ic4. White is clearly better. Also 14 . . . .!e7 1 5 . lLl d6t ixd6 1 6.exd6 is just clearly better for White.) 1 5 . lLl d6t \tle7 1 6.�b3 Wfa4 17.Wfb6 Eihc8 1 8 .lLlxc8t and in the game Radjabov - Dominguez, Tripoli 2004, White won within a few moves. 9.0-0-0 �bd7

9 . . . 0-0 is an old sideline. Now White should play 1 O .g4. (This move seems to me more in the spirit of the position. 10 .id3 lLl c6 l 1 . lLlxc6 bxc6 12 .�g3 was played in Unzicker - Fischer, Buenos Aires 1 960, after 12 . . . e5 Black could have achieved a nice game.) 1 0 . . . lLlc6 l 1 .ih4! (of course White keeps this bishop now) l 1 . . .lLlxd4 12 . Eixd4 b5 13 .g5 and White later won, Hector - Evertsson, Stockholm 1 999.


10.g4 b5 u.J.xf6 ttlxf6 l l . . .ixf6 12.g5 ( 12 .ixb5 0-0 ( 12 . . .

axb5? This i s the wrong reply. 1 3 .4:Jdxb5 'lMfbB 14.4:Jxd6t 1!;>f8 1 5 .e5 and White is winning.) 13 .ixd7 ixd7 Black has fine compensation for the pawn.) 1 2 . . . ixd4 ( 12 . . . ie7 transposes to the main line) 1 3Jhd4 0-O 14 .ih3. White was better in Luther - Ardeleanu, Linares 1 99B.

1 1 . . .gxf6 is another option here. Since White has played g2-g4 it is not so easy for him to attack e6. 1 2.5 4:Je5 13 .'lMfg3 (this is better than the more common 13 .Wih3) 1 3 . . . id7 14 .ih3 4:J c6 1 5 . 4:J xc6 ixc6 1 6.fxe6 fxe6 17.g5 and White was much better in Guseinov - Makoll, Turkey 2004. 12.g5 ttld7 13.f5

Here many moves have been tried. For example: a2-a3, if1-h3, h2-h4, etc. I think only the text offers White attacking chances.

1 3 . . . ixg5t. Taking this pawn is the principled reaction. 14.l!;>b 1 4:Je5 1 5 . 'lMfh5 if6! (This move is better than the more common moves 1 5 . . . 'lMfe7 and 1 5 . . . 'lMfdB: 1 5 . . . 'lMfdB 16 .h4! if6 17.fxe6 0-0 1 B .ih3. White was better in Luther - Vink, Wijk aan Zee 200 1 .) 1 6.4:Jxe6 ( 16.fxe6 Leaving the bishop cB alive later causes White some problems. 1 6 . . . g6 17.4:Jd5 WidB 1 B .exf7t 1!;>xf7 and suddenly White's queen is in trouble.) 16 .. . ixe6 17.fxe6 and the position needs further investigation. 14.£6 gxf6 15.gxf6 .tfs 16J�tgl h5 17J�el!

This move was found by GM Peter Enders. While working on the Fritz opening book I entered this move as a recommendation for the computer program. Anyway, some time later I had the chance to play this move in a tournament game.

17.a3 is another possible move, but I prefer the text.

1 7. l::1g7 This funny exchange sacrifice was played a couple of times in the late BOs and early 90s. White had some nice victories but today's computer analysis proved it is incorrect. 17 ... .id7

17 . . . b4 does not work here. 1 B .4:Jd5 ! This is the main idea behind 17. l::1 e 1 .

1 7 . . . ib7?! 1 B .ih3 . . . 0-0-0 1 9 .4:Jd5 with a strong initiative. 18.c.i?bl

A useful waiting move. White wants

26 The Najdorf

to see where Black's king is going before committing himself to one specific line of action. 18 ... 0--0-0?!

This must be wrong since now White is in control of the game. 1 8 ... b4 was called for. 1 9.tilce2. There is no need for til d5 here as Black cannot bring his bishop to b7 anymore. White is better. 19.9g5

Not every game is won by a great attack. White just wants to collect the h-pawn. 19 ... h4 2oJ�Ml!?

The rook has done its job on e 1 and now defending the til d4 is necessary to avoid counterplay. 2o .. .cbb8

21 .a3± Taking the last resource (b5-b4) out of

Black's play. 21 . .. 'ffb6 22Jl�h5!

Finally the h-pawn is lost. 22 .. .j,b6

Black wants to bring his passive bishop into the game, but there is no way it will reach a comfortable square. 23J�xb4 Ag5 24.gg4 Ah4?!

This is not leading anywhere. 24 ... .th6 was better. 25.J.g2

Preparing e4-e5. 25 ... �a7 26.gg7 gdf8 27.e5!

Now it comes. Black is lost. 27 ... d5 28.'fff4

The conclusion comes from the other side. ih4 is trapped. 28 ... c!Lla4 29.c!Lla2!

Preventing any counterplay! 29 .. . gh5 30.AS ghh8 31 .Ae2 'ffa5

The last try. 32.gg4 .leI 33.c!Llb3

Finally Black's bishop is lost. 1-0

Recently a friend asked me what I recommend against the Gelfand-line. I was surprised but I knew which line he meant. GM Boris Gelfand has played many famous games with the Najdorf Defence and his usual set-up against 6.ig5 involves not developing the .tfS , but looking for a quick mobilisation of his queenside pieces. The move 1 1 . . . \Wb6 became popular after he used it against GM Shabalov. We will look at this game now.


Game 8 Shabalov - Gelfand Bermuda 2004

1 .e4 e5 2.�a d6 3.d4 cx:d4 4.�xd4 �f6 5.�c3 a6 6..tg5 e6 7.£4 �bd7 8.Ba Be7 9.0-0-0 b5

Black plays without ifS-e7 and focuses on quick development of the queenside. lo..td3 ib7

1 0 . . . b4 1 1 .1L1dS exdS 12 J �hel 1eads to a previously mentioned game. 1 l .ghe1 Bb6

For 1 1 . . .ie7 see Kotronias - Lesiege. 1 1 . . .�. This has been played a few

times. I think the black king is misplaced. Black's position is losing all its dynamics. 1 2.5 eS 1 3 .1L1b3 ie7

1 4.a4! . Exacdy in this moment White weakens all the light squares in Black's territory. 14 . . . b4 I s .ixf6! lLlxf6 16 .1L1dS White dominates the position. 16 . . . 1L1 xdS 1 7.exdS as 1 8J !e4 @b8 19 J �k4 'iNb6

20.ie4 with a clear advantage for White in Kotronias - Jobava, Batumi 2002. On 20 .. Jk8 then 2Uk6 is a standard exchange sacrifice. 2 1 . . .ixc6 22.dxc6 !=(hd8 23.@b l dS 24.ixdS and Black will not find happiness in this game. 12.�d5

If White wants to prove something in this position he has to jump forward. Any retreat, such as lLld4-b3, gives Black an easy life. 12 ••• Bxd4

1 2 . . . exdS 1 3 .1L1c6! This jump from the knight causes Black a lot of trouble. 13..bf6 gxf6 14.1xb5 Be5

15.h4? Here Shabalov goes wrong. White wins

the queen, but in return Black gets three minor pieces and lots of good squares.

I S .1L1xf6t @d8 1 6.1L1xd7 'iNxbS 17.1L1xfS ( 17.'iNc3 does not offer anything after 1 7 . . . @xd7 1 8 .'iNxh8 'iNhS) 17 . . J:J:xfS 1 8 .'iNa3! ( 1 8.!=(xd6t is also

28 The Najdorf

possible, but the text is better. 1 8 . . . l!;>c7 1 9J:ledl Eiad8 20.'�c3t i.c6 2 1 .e5.) 18 . . . l!;>e8 19 .'�xd6 Eic8 20.b3. With three pawns for the bishop, White keeps an advantage. In the coming endgame he plays c2-c4 and places his king on c3 . l5 ••• YlYxb5 l6.�e7t rt;e7 l7.�xb5 axb5 l8.YlYh5 gxa2

1 8 . . . i.c6 is a much safer reply. 19.YlYxb5 �6 20.e5 £Xe5?

Better was 20 . . . Eic8, which secures Black a big advantage. 2 1 .exd6t I!;>d8 22.'�·d3 i.xf4t 23.l!;>bl Eicxc2 24.'�xc2 Eixc2 25 .l!;>xc2 i.xg2 black is winning. 21 .YlYxb7 hf4t 22.<bbl ghaS 23.g3 g2a7 24.YlYc6 ga6 25.YlYc3

I have the impression that White should not have lost this position. On the other hand he certainly has no advantage either. 25 ••• �6 26J:le4 d5 27.gh4 d4 28.YlYb2 J.g5 29.gxh7 �f6 30.M �d5 31 .gd3 J.d2 32.gxd2 �c3t 33.<bcl galt 34.YlYxal gxalt 35.<bb2 gblt 36.<ba3 e4 37.gn e3 38.gfxf7t <bd6 39J�ld7t <be6 40.ge7t <bd5 41 .ged7t <be4 42.gde7 <ba 43.gxe6 <be2 44.gd7 <bd2!

A nice manoeuvre: Black's king joins the mating attack. 45.gxe3 <bxe3 46.gf'7 �d5 47.b5 <bd2 0-1

The sacrifice on b5 is a well-known idea in the NajdorE Instead of developing the bishop from f1 White immediately sacs it

for two pawns. The lead in development and the possible capture of a third pawn (usually the one on d6) makes it possible. Here this sacrifice is connected with e4-e5 after which many tactical lines arise. I think there is plenty of scope for improvements in this line. However even the typical endings in this line (queenside passed pawns for White vs. a piece, or passed pawn + rook vs. 2 minor pieces) are not clear. I can only give the reader a brief summary of this line. I recommend further study before trying it in a tournament game.

Game 9 Timman - Gelfand Wijk aan Zee 2002

1 .e4 e5 2.�a d6 3.d4 cxd4 4.�xd4 �f6 5.�c3 a6 6.J.g5 e6 7.f4 �bd7 8.YlYa YlYe7 9.0-0-0 b5

Black plays without i.f8-e7 and focuses


on quick development of the queenside. 10.bb5

This move was for a long time considered the main reply to Black's set-up.

1 0.e5. The most forcing reply. Unfortunately, nowadays most of these purely tactical lines have been analysed till the end. This line is a forced draw, where neither side can avoid it. 1 O . . . .tb7 I I .'lWh3 dxe5 12 .ttJxe6! fxe6 1 3 .'lWxe6t .te7 14 . .txf6 gxf6 1 5 . .te2 h5 16. ttJ d5 .txd5 1 nhd5 ttJ c5 ( 17 . . . ttJb6 1 B . .txh5t :B:xh5 19 .'lWgBt=) I B .'?Nf5 '?Nc6 1 9.'?Ng6t �fB 20. :B:hdl 'lWeB 2 1 .'?Nf5 '?NcB 22.'?Ng6 '?Ne6 23 . .txh5 '?NgB 24. :B:dBt :B:xdB . Draw agreed in van der Wiel - Kasparov, Amsterdam 1 99 1 . 24 . . . :B:xdB 25. :B:xdBt .txdB 26.'?NeBt �g7 27.'?Ng6t= is the perpetual behind the draw.

1 0 . .txf6 ttJxf6 ( 1 0 . . . gxf6 1 1 .5 gives White a good game) I 1 .e5 .tb7 12.'lWe3 (This is a recent discovery from GM J .Hector. 12 .'lWh3 dxe5 1 3 .ttJcxb5 'lWb6 ( 1 3 . . . axb5? 14 . .txb5t �e7 1 5 .fxe5 ttJd5 1 6. 'lWh4t f6 17.exf6t gxf6 I B . :B:he l e5 1 9 . ttJ f5t �f7 20 . .tc4 and White wins his material back and gains a winning advantage.) 14.fxe5 ttJ e4. Black has fine compensation for his sacrificed pawn.) 12 .. . dxe5 1 3 .ttJcxb5 '?Nb6 14.fxe5 ttJd5 1 5 .'lWg3 O-O-O? ( 1 5 . . . axb5 reminds me of the old saying ''A sacrifice can only be refuted by accepting it." 16 . .txb5t �dB 17.a4 ( 1 7.c4.This wins back the piece, but Black has easy play afterwards. 17 . . . :B:xa2 I B.cxd5 .txd5 19 . � b 1 .tc5 20. ttJ b3

:B:xb2t 2 1 . �xb2 'lWxb5.) 17 . . . �cB I B . :B:hf1 . White has some play here, but I recommend that the reader checks this position with his computer before testing it over the board.) 1 6. ttJ a3 .txa3 17.'?Nxa3 ttJ b4 I B . .tc4 :B:xd4 19Jhd4 '?Nxd4 20.'?Nxb4 :B:dB 2 1 .'lWc3 '?Nxc3 22.bxc3 White is better, Hector - van der Stricht, Plovdiv 2003. 10 ... axb5 1 1 .c!L!dxb5 VNb8 12.e5 ga5

12 . . . .tb7? For a long time this was considered to be a good alternative to the main line, but recent computer analysis showed that White gets a decisive advantage. 1 3 .'lWe2 dxe5 14.'?Nc4 .te7 1 5 . ttJ c7t �f8 16. :B:xd7 ttJxd7 17 J!dl .txg5 I B .fxg5 .tcB 1 9. ttJ 3b5 g6. So far everything was played in Kengis -Dvoirys, USSR 1 9B2.

Now White should play 20. :B:fl �gB 2 1 . ttJ d6! and the computers show that White has a clear advantage. The mating threat after ttJxf7 forces Black to make a

30 The Najdorf

serious concession. 2 1 . . .l'ha2 22.'\Mf xa2

Wixc7 23. ttlxf7 ttl c5 24. ttlh6t 'it>g7 25 . E:f7t Wixf7 26. ttlxf7 'it>xf7 27.'�'a7t ttl d7 2B.c4 Black lacks coordination and is helpless against the advance of White's pawns. 13.exf6 gxf6

1 3 . . . E:xb5 14 . ttlxb5 Wixb5 1 5 .Wic3+-. 14 • .th6!

Rather funny, the bishop sacs itself 14 •. ..bh6 15.tLJxd6t rtJe7 16.rtJb1

There was the threat of ttl d7-e5. 16 ••• Eld8

1 6 . . . ttl b6!? 1 7. ttl cb5 E:xb5 I B . ttl xb5 �xf4 19 .'�' c3 ttl d7 20.Wia3t 'it>eB 2 1 . ttl d6t .txd6 22. E:xd6 E:gB 23 .g3 E:g5 24. E:hd l . This position requires further investigation, but I think White has the better play here. 17.Elhe1 tLJ b6

17 . . . �g7 I B .Wic6 and, thanks to the threats of ttlxf7 and ttlxcB followed by ttl d5t, White wins material. 18.tLJcb5

18 • • J�xb5 This move does not seem to be the

best. I B . . . E:d7 19 .Wic6 'it>f8 20. ttlxcB

E:xdl t 2 1 . E:xdl ttlxcB 22. E:dBt 'it>g7 23.'�'eB 'it>g6 24.fSt 'it>xf5 25 . ttl d4t 'it>e4 26. Wixf7. White is winning.

Half a year later Gelfand played I B . . . .ta6 1 9. Wic3 ( 19 . ttl fSt. This move leads to a draw. 1 9 . . . 'it>f8 20.\Mrc3! . A forced move again, as is the whole line actually! 20 . . . E:xb5 2 1 . \MrxfG E:xb2t! . This surprising capture saves the day for Black. 22.\Mrxb2 ttl d5 23. E:xd5! . Another forced blow continues the series. 23 . . . Wixb2t 24. 'it>xb2 �g7t! . This check secures Black the draw. 25 . ttlxg7 E:xd5 26. ttlxe6t fxe6 27. E:xe6 Y2-Y2 Lutz - Gelfand, Dortmund 2002.) 1 9 . . . E:xb5 20. ttlxb5 �xb5 2 1 .\Mrb4t E:d6 22.Wixb5 �xf4. So far as in Kantsler - Kuporosov, Sochi 1 979. Now White should have played 23.a4 Wi c7 24. E:xd6 It is not easy to assess this position, but the passers on the queenside should give White better chances. 19.tLJxb5 Elxd1 t!?

19 . . . ttl c4 forces a difficult ending. I recommend that the reader study this kind of ending before testing it over the board. 20.Wib3 ttl d2t 2 1 . E:xd2 E:xd2 22. \Mrb4 t E:d6 23.g3! . Winning a tempo. 23 . . . �d7 24.\Mrxd6t Wixd6 25 . ttl xd6 'it>xd6 26. E:dl t! 'it>c7 27. E:fl . After stopping e6-e5 for some time, White shall advance his queenside pawns. I cannot


come up with a final conclusion, but my feeling is that White's position is fine. 20.gxdl hf4 21.g3

Again we see the same kind of ending arising: rook + 3 passed pawns vs. 2 minor pieces + a central passed pawn. 21 . .. J.e5 22.Y!Ya3t We8 23.�d6t!

It is very much in White's favour here to exchange pieces. 23 ... J.Xli6 24.Y!Yxd6 Y!Yxd6 25J:�xd6 �d5 26.c4 We7??

A horrible blunder by Gelfand, but even after other moves White keeps an advantage. Once the queens are exchanged White can safely advance his queenside pawns.

27.gc6+- J.b7 28.cxd5 J.xc6 29.dxc6 Wd6 30.g4

The pawn ending is an easy win. 1-0

Thanks to the following game the move 1 0 . .!d3 enjoyed a revival. GM Kotronias

played a great novelty after which Black faces a very difficult defence. In many variations White has long lasting compensation due to the unfortunately placed black pieces. This game is a nice example to study when White is ready to sacrifice his c3 knight on d5. It is very important that the bishop on f8 remains weak and does not enter the game.

Game 10 Kotronias - Lesiege Montreal 2002

1 .e4 c5 2.�f3 d6 3.d4 cxd4 4.�xd4 �f6 5.�c3 a6 6..tgs e6 7.£4 �bd7

7 . . . h6?! From today's knowledge I can say that Black should not play h7-h6 at any stage of this opening. 8 . .!h4. Now there are two main lines:

a) 8 . . . .!e7 9 .�f3 �c7 (9 . . . g5 . Black wants to fight for the e5 square. But even if he manages to place one of his knights

32 The Najdorf

there it is not clear ifhe has achieved much. 1 0.fxg5 lLl fd7 l 1 . lLlxe6!. White is going for a direct refutation. 1 1 . . . fxe6 12 . �h5t mfB 13 .!b5! This was first seen in the 50s. The most famous game is Gligoric - Fischer, Portoroz 1 955 . Much has been written about this game and finally computer analysis proved that White is winning here. In the original game the American future World Champion saved the draw. A recent example shows how dangerous Black's position is: 1 3 . . J��h7 14 .0-0t mgS 1 5 .g6 �g7 1 6. �f7 !xh4 17.'1&xh6 �xf7 I S .gxf7t mxf7 19 . �f1 t !f6 20.e5 dxe5 2 1 . lLl e4 axb5 22.1�'h7t mfB 23. �hSt me7 24. �h7t mfB 25 . lLlxf6 �b6t?? (25 . . . �xf6 26.�hSt mf7 27. �xf6t lLlxf6 2S.�xcS lLl fd7t and the fight continues) 26. mhl lLlxf6 27. �xf6t meS 2S .�f7 1-0 Naiditsch - Enders, Hockendorf 2004.) 1 0.0-0-0 lLl bd7 1 1 .!d3 g5 (The same idea as mentioned above. At least Black is not getting checkmated immediately. 1 1 . . . b5? This runs into a direct refutation. 1 2.e5 !b7 1 3 .lLlxe6 fxe6 14.!g6t Here is the difference from positions without h6 and !h4 included: White has this check. 14 . . . mfB 1 5 .exf6 !xf3 ( l 5 . . . !xf6 16 .�h3 Black has too many weaknesses and no compensation for them in return) 1 6.fxe7t mgS 17.gxf3 with a big material advantage.) 1 2.fxg5 lLle5 1 3 . �e2 lLl fg4 14. lLl f3 lLlxf3 1 5 .�xf3 hxg5 ( l 5 . . . lLl e5 16 .�h5 lLlg6 17.!g3 hxg5 I S .�f3 !d7 White is better, but Black can definitely

play this position. Luther - Ginsburg, Germany 2004.) 16 .!g3 lLl e5 1 7.!xe5 dxe5 I S . �dfl ! �h7 1 9.h4! White's attack is coming too fast for Black to set up a defence. Luther - Ott, Hockendorf 2004.

b) S . . . lLlbd7 9 .�f3 e5 A relatively new and surprising idea. 1 0. lLl fS ( l 0.fxe5? This is the wrong reaction. 1 0 . . . lLlxe5 1 1 . � dl !e7 12.!e2 lLlxe4 13 . .txe7 lLlxc3 14 .!xdS lLlxdl 1 5 .!c7 lLlxb2 1 6.!xd6 lLlbc4 with a decisive advantage for Black in Radjabov - Dominguez, Tripoli 2004.) 10 . . . exf4 1 1 . �xf4 g5 12 .lLlxd6t !xd6 13 .�xd6 gxh4 14.e5 �e7 1 5 .�xe7t mxe7 1 6.exf6t. The tactics are finished and White has a better ending. s.We Wc7 9.0--0-0 b5 lo.id3 ib7

1 O . . . b4 l 1 . lLl d5 exd5 1 2. �he l with transposition to the lines mentioned below. 1 l .B:he1 i.e7 12.Wg3 b4 13.�d5 ad5

14.exd5! This game of V. Kotronias changed

the evaluation of the line. Before it the theory was 14.eS dxeS I S .fxeS lLlhS 1 6.�h4 .!xgst 17.�xgS g6 1 8 .e6 lLlcS 1 9 .exf7t c;t>xf7 20JW t c;t>g8 2 1 . lLl fS lLl e6 22. lLlh6t with a perpetual, Luther Nielsen, Malmo 2002. There were many other lines analysed and games played, but theoretically they belong to the past. 14 ... <;bd8 15.�c6t .hc6 16.dxc6

Here Black has a large choice. Most of the variations I give below I analysed with my computer. Since soft- and hardware develop fast and this is a highly tactical position, I recommend that the reader check his favourite lines. I have played this position twice (vs. GMs Shneider and Dvoirys.) and I think that in practice the piece sacrifice is fully justified. 16 ... �c5

1 6 . . . lLlb6 17 . .!h4 .!f8 ( l7 . . J !g8 1 8 .�xh7 l"ih8 1 9 .�e4 dS ( l 9 . . . lLlxe4?

20. �xg7 and Black is lost) 20.�xf6 �xf6 2 1 ..!xdS lLlxdS 22. l"ixdSt with a strong attack.) 1 8 .�xf6t gxf6 19 .�e4 and White obtained the typical compensation and has good practical chances.

1 6 . . . YMxc6 The most principled reply. Black eliminates the dangerous passer. 17 .�xf6 leaves Black with a difficult choice. 1 7 . . . lLlxf6 ( l7 . . . �xf6 1 8 .�e4 YM a4 Black has no big choice here. (The alternative 1 8 . . . YMc8 19 .�xa8 �xa8 20. l"ixd6 l"ie8 2 1 . l"ixd7t c;t>xd7 22.�d3t c;t>c7 23.�c4t c;t>b6 24.�xb4t c;t>c7 2S .�cSt c;t>b7 26.YMdSt allows White an instant win) 19 .�xa8 �xa2 20. l"ixd6 �xb2t 2 1 . c;t>dl �bl t 22. c;t>e2 YMxc2t 23. c;t>fl! (23 . c;t>f3?? In my game against Dvoirys I chose the wrong square for the king. Now Black could save the game. 23 . . . l"ie8 24. l"ixe8t c;t>xe8 2S.�c6 c;t>f8! 26 . .!xd7 c;t>e7 27. l"ixf6 gxf6 28.�g4 b3 29.YMelt c;t>f8 30.YMaS= Luther - Dvoirys, Austria 2003 .) 23 . . . YMc4t 24. l"id3 l"ie8 2S . l"iedl l"ie7 26.YMf3 and White is winning.) 1 8 .�xg7 l"ig8 19 .�xf7 �e8 20.YMc4 (20.�b3?! ' This move is weaker than the text, but White is still better, Luther - Shneider, Istanbul 2003.) 20 . . . l"ic8 2 1 . YMxb4 YMc6 22.g3 and White is clearly better. Besides the three pawns he has for the knight, Black's king is in extreme danger.

1 6 . . . h6 This move also does not change the evaluation of the position, White picks up another pawn in compensation for the sacrificed knight and is continuing

34 The Najdorf

the attack. 1 7.!h4 gS I S .fxgS hxgS 19 .1!;lfxgS lLlcS 20.i.f5 White is clearly better here. 17 . .th4!

17 . . . .ifB This makes it easier for White. More

trouble is: 17 . . . g6 I S .f5. This line does not change

the final conclusion either: White keeps attacking.

17 . . J �gS 1 8 .!c4! ? lLl fe4 ( 1 S . . . 1L1ce4 1 9.1!;lff3 and White controls the game) 19 .1!;lfg4 i.f6. A typical computer move, as my friend Jacob Aagaard pointed out.

a) White has easier play after: 1 9 . . . !xh4 20.�xh4t 1Ll f6 (20 . . . �e7? 2 1 . �xd6t Losing d6 is bad news for Black. White is winning.) 2 1 .g4 @cS 22.gS and White is overrunning Black's strongholds.

b) 1 9 . . . �a7 20.i.dS i.xh4 2 1 .�xh4t 1Ll f6 22.i.f3 with the already mentioned ideaofg2-g4-gS. 20.i.dS ixh42 1 .1!;lfxh4t 1Ll f6 22.g4 again with initiative.

However, just as the book was finished and ready to go into print a new masterpiece of invention was revealed by our Greek hero Kotronias:

l S .i.xh7! ! . This very surprising move does not tempt the computers, but all will become clear soon enough. l S . . . �hS 1 9.�xg7 �xh7 20.�xf6 �xh4 (20 . . . ixf6?? 2 1 . i.xf6t @cS 22. �eSt is bad style) 2 1 .1!;lfxf7 �hS (2 1 . . . �g4 22. �dS! with the idea of 23 . �hS and Black has no defence) 22. �eS lLl a4 23 . �e6 lLlcS (the computer move does not really work here: 23 . . . 1L1b6 24. �de 1 lLlcs 2S .�g7 �eS 26.f5 �xc6 27.f6±) 24. �eS lLl a4 2S .�e3 �a7 (Black cannot pretend to repeat the line, as after 2S . . . 1L1cS 26. �de 1 he has nothing better than 26 . . . �xc6± with a bad endgame. 26 . . . �eS 27. �xe7!+-.) 26. �de1 �eS (A possible alternative is 26 . . . �xc6 27. �xe7 �xe7 2S.1!;lfxe7t @c8 29.h4;±; but White certainly still has all the chances, and should not be unhappy about reaching this kind of position.) 27.f5 �xc6 2S.f6 1!;lfxg2? After this the white attack wins without any problems. (2s . . . 1L1 cS 29.fxe7t @cS and 2S . . . �c7 29. � l e2 are probably both only slightly better for White and certainly what Black should have opted for.) 29.fxe7t @d7 30. �g3 �aS 3 1 . �g6 1L1cS 32. �d1 1L1b7 33.�e6t @c7 34. �g4 as 3S .�c4t 1-0 Kotronias - Shneider, Korinthos 2004. 18.J.c4

It is this kind of position White is


basically aiming for. Sooner or later he will take on f6 and ruin Black's pawn structure. 1fS will be locked in and White penetrates Black's position via the e-line. The main game is a perfect example of transferring White's strategy into a win. IS •. J�a7 19 . .idS as 20J:�e3 'ilYc8 21 .gdel 'ilYf5 22.hf6t gxf6 23.geSt �e7 24.'ilYa

White has no direct threats, but he controls the game. Black suffers from the lack of coordination between his pieces. 24 •.. hS 2S.�bl .ig7?

Making White's task easy, but staying passive is not what one wants to do in Black's position. 26J�xhS .ixhs 27.hf7 �b6 2S.J.xhS a4 29.i.g4 'ilYh7

30.e7! This pawn has done its job. Now White

opens the lines to the black king. 30 •• J!xe7 31 .'ilYdS f5 32.'ilYxd6t ge6 33.'ilYdSt 'ilYe7 34.'ilYxhS fxg4 3S.'ilYd4

�bS 36.b3! axb3 37.axb3'ilYd6 38.'ilYe4t �aS 39.:aeS

Still Black is completely tied up. 39 ••• 'ilYdlt 40.�a2 'ilYd6 41 .:adS 'ilYe7 42.f5 �b6 43.'ilYxb4t �a7 44.'ilYd4

A great game! 1-0

Many players consider the Poisoned Pawn variation as the main reply against the ig5 system. White has no good way to avoid losing the b2 pawn so he is committed to attacking Black at all costs. In a possible arising ending White would not only be a pawn down but also have a weak pawn structure on the queenside. On the other hand White gets a large advantage in development and the threats can be really dangerous. In this game GM Short came very close to beating the World Champion.

Game 1 1 Short - Kasparov Riga 1 995

1.e4 eS 2.�a d6 3.d4 cxd4 4.�xd4 �f6 S.�c3 a6 6 . .ig5 e6 7.£4 'ilYb6

7 ... h6 8.ih4 �b6. The combination of these two lines is not recommendable for Black. 9.a3! with the idea of in. Of course 9 .. .'�'xb2 10.�a4 wins the queen. S.'ilYd2

The only reasonable way to play the position.

36 The Najdorf

B.a3 is just a tricky sideline. B . . . lU c6 9. lUb3 Wfe3t! (9 . . . i.e7 1O .Wff3 h6 1 1 .i.xf6 i.xf6 12.0-0-0 Wfc7 13 .h4 i.d7 14.g4 0-0-0 I S .i.g2 c;!{bB 1 6.gS i.e7 17 .Wfe3 i.cB was played in Luther - Dominguez, Havana 200 1 . The position is unclear.) 10 . Wfe2 Wfxe2t After exchanging queens Black has no problems at all.

B .lUb3 i.e7 (B . . . �e3t 9.�e2 �xe2t is a very sound continuation for Black. I think the only reason why it is played so rarely is that the usual Najdorf player does not want to enter an even ending, but instead prefers to attack.) 9. � f3 lU bd7 10 .0-0-0 �c7 1 1 .i.d3 bS. Black has reached his normal set-up, but White's knight is no longer on d4 but on b3, so all the nice tactics do not work anymore. 1 2.a3 White would rather avoid this move but after 12 . �hel b4 White has to remove the knight and Black gets very comfortable play. 1 2 . . . �bB! Black's attack comes faster now. 8 ... '§'xb2

B . . . lU c6. This sideline has been seen often in tournament games recently. Black decides not to take on b2 but transfers the game into a Rauzer-like set-up. 9 .i.xf6! (Only with this move can White fight for an advantage. This capture is a standard reply if Black has to take back with the pawn. Instead 9.0-0-0 �xd4 1O .Wfxd4 lUxd4 1 1 . �xd4 i.d7 with an even ending.) 9 . . . gxf6 1 O .lUb3 i.d7 1 1 . i.e2 hS If Black does not play this

move White will place his bishop on hS. 12.0-0-0 0-0-0 13 .c;!{bl c;!{bB 14 .h4! . A strong idea. White fixes hS as a weakness and prepares to bring the h I rook into the game. 14 . . . lUaS I S . �h3 �cB 16 .lUxaS �xaS 17. �d3 with advantage for White in Sulskis - Loginov, Vilnius 1 997. 9.tLlb3

9 .�bl This move leads in many lines to a forced draw. 9 . . . �a3 1 O.f5 lUc6 1 1 .fxe6 fxe6 12 . lUxc6 bxc6 13 .eS ( 13 .i.e2 This is a famous drawing line which has been seen many times in tournament games: 1 3 . . . i.e7 14.0-0 0-0 I S . �b3 �cst 16 .i.e3 �eS 17.i.d4 �aS I B .i.b6=) 13 . . . dxeS 14.i.xf6 gxf6 I S . lU e4 �xa2! (1 S . . . i.e7? This move brings Black trouble. 1 6.i.e2 hS

17 . �fl ! f5 1 B . �f3 vgxa2 19 . �fb3! . A great idea of the White player! Black's queen is cut off for a long time; meanwhile White focuses on Black's king. 1 9 . . . fxe4 20.�c3 i.dB (20 . . . 0-0 2 1 . �al i.b4 22. �xb4


�d5 23.i.xh5 looks promIsmg for White.) 2 1 .�xc6t i.d7 22.�xe4 �f7 (22 . . . �a5t 23. �fl o-ot 24. Eif3 i.f6 25.i.d3 �f7 26J':lb7 and White's attack continues.) 23. �fl . Further analysis is required but I have great faith in White's position, Gubajdullin - Biriukov, St Petersburg 2003.) 16 . Eid1 i.e7 17 .i.e2 0-0 1 B .0-0 Eia7 1 9 .Eif3 �hB 20. Eig3 Eid7 2 1 . Wlh6 Eif7 22. �h5 Eixd1 t 23.i.xd1 Wl a5 24. �fl Wl dB 25. �xf7 �xdlt and soon the game ended in a perpetual, Vallejo Pons - Kasparov, Moscow 2004.

9 . . . Wla3 9 . . . lD bd7 10 . i.xf6 gxf6 ( 10 . . . lD xf6?

1 1 .a4 traps Black's queen) 1 1 .i.e2 leads to the main line.

9 . . . lDc6. This move disappeared from tournament practice because it leads to a better ending for White. 1 0.i.xf6 gxf6 1 1 . lDa4 Wla3 1 2. lD b6 EibB 1 3 .lDc4 �a4 14.a3! ( 14. �f2. This is an old line, but

after 14 . . . i.e7 1 5 . lDxd6t �f8 White has nothing.) 14 . . . b5 1 5 .lDxd6t i.xd6 16 .�xd6 Wlxe4t 17.i.e2 �d5 There is nothing better then this.

a) 17 . . . �xg2? 1 B .0-0-0 i.b7 19 .�d7t �f8 20.Wld6t �eB 2 1 . i.h5 leads to a winning position for White.

b) 17 . . . i.b7. This move does not solve Black's problems either. 1 B .lD c5 ( 1 B .O-O-O EidB and Black is doing fine) 1 B . . . Wld4 19 . Eid1 Wlxd6 20. Eixd6 lD dB 2 1 . lDxb7 lDxb7 22. Eixa6 with a better ending for White. 1 B .�xd5 exd5 1 9 .0-0-0 lD e7 ( 1 9 . . . i.e6 20.g4 f5 2 1 . Eihe 1 lDe7 22.i.d3 �fB (22 . . . Eib6 23.gxf5 (23 .Eid2 �dB 24.gxf5 i.cB 25. Eide2.) 23 ... i.cB) 23.gxf5 lDxf5 24. Eie5 lD e7 25 . lDc5 Eib6 26.f5 Eic6 27. lDxe6t fxe6 2B.fxe6 �eB Y2-YZ Rogers - van der Sterren, Hertogenbosch 1 999) .

Now White can try either 20.i.f3 or 20. Eihe 1 . Let us look at 20.Eihe1 first:

a) 20 . . . 0-0 2 1 . i.xb5 was played in

38 The Najdorf

Kotronias - Ftacnik, Pula 1 997, White is better. (2 1 .if3 ie6 leads to the above mentioned line.) .

b) 20 . . . E:b6 2 1 .if3 E:e6 22. tLl d4 E:xe l 23. E:xe l c,:!;d8

c) 20 . . . E:b6. This reply saves Black. 2 1 .if3 E:e6 22. tLld4 E:xel 23.E:xe l c,:!;d8=. After exchanging one pair of rooks, Black has equality.

After 20.if3 the normal reply is 20 . . . ie6 and only now 2 1 . E:he1 O-O! (2 1 . . . E:g8 22. tLl d4 E:b6 23.f5 gives White a pleasant advantage) . Now 22.g4! is best. Other moves promise less:

a) 22. tLl c5 E:b6 23. tLld7 ixd7 24. E:xe7 ie6 25 .ixd5 ixd5 26. E:xd5 E:c8 with equality.

b) 22. tLl d4 E:b6 23.g4 f5 24.gxf5 tLlxfS 25 .tLlxf5 ixfS 26. E:xd5 ie6 and again the game is even.

c) After 22.g4! play can continue 22 . . . f5 23.gxf5 tLlxfS 24.ixd5 E:be8! (24 . . . E:b6? Despite the reduced material it is not too late for Black to blunder. 25 . tLl c5! ic8 26.ib3 !± with much better play for White) 25 .ib7!? (25 .ixe6 E!:xe6 �-� Adams - Gelfand, Wijk aan Zee (3) 1 994) 25 . . . ic8 (25 . . . E!:b8? 26. tLlc5 !± . 25 . . . ixb3 26.cxb3 E:xel 27. E:xe 1 tLl dM .To win this position with White requires good technique, but defending Black's side is no fun either.) 26.ixc8 E:xc8 27. E:e5 tLlh4 28J:(g5t c,:!;h8 29. E!:d6 and White keeps some advantage. lo.ixf6 gxf6 ll .i.e2

1 l . . . h5 1 1 . . . tLl c6. This is another standard

reply. Black wants to bring his rook to c8 as quickly as possible. 1 2.0-0 id7 ( 12 . . . ig7?! 1 3 .E!:f3 0-0. Black has chosen a very dangerous set-up. White's pieces will target Black's king very rapidly. 14 .c,:!;hl E:d8 1 5 . E:g3 d5 1 6.exd5 f5 17 .E:dl �f8 1 8 .d6 c,:!;h8 1 9 .tLla4 with a large advantage in Kasimdzhanov - Polgar, Moscow 2002.) 1 3 . c,:!;hl E!:c8 ( 1 3 . . . h5 14 . tLldl E:c8 1 5 . tLl e3 �b4 1 6.c3 �xe4 17 . tLlc4 tLld8 1 8 . tLl d4 h4 1 9 .if3 �h7 20. tLl b6 E:c5 2 1 . tLlxd7 c,:!;xd7 22. E:ab l h3 23.g3 E:c7 24.f5. This was played in Kotronias - Sasikiran, Moscow 2004. For the two sacrificed pawns White has a strong initiative. Black's pieces are bad placed, especially his king and queen. White managed to win this game later, but it requires more analysis to come to a final conclusion about the position.) 14 .ih5?! (Too early,


White should wait with this move until Black has played iLfS-e7. Better is 14J'!f3 to protect c3 and prepare some action on the kingside. 14J'!adl iLe7 1 5 .iLh5 l.'!fS 1 6.l.'!f3 �d8 17 . l.'!d3 l.'!c7 1 8 .lLlb l 'I1;!ia4 1 9 .1Llc3 �a3 20. lLl b l 'I1;!ia4 2 1 . lLl c3 � a3 was played in Luther - Georgiev, France 2003. All White's pieces are in good positions, but Black is rather solid too.) 14 . . . iLg7! 1 5 .l.'!f3 0-0. Now this is a good idea. The iLh5 is misplaced and slows down White's attack. 1 6. l.'!af1 lLla5 17.5 lLlc4 1 8 . �f4 lLl e5 and Black was better in Kasimdzhanov - Sadvakasov, Skanderborg 2003

l 1 . . .iLg7 Black should not play this move so early. The bishop on g7 can be attacked later by l.'!g3 . 1 2.0-0 �b4 13 . l.'!f3! lLl d7 14 .l.'!dl 'I1;!ib6t 1 5 . �h l 'I1;!ic7 1 6. l.'!g3 iLf8 17 .iLh5 lLlc5 1 8 .lLlxc5 'I1;!ixc5 1 9 .e5! . After this standard move White gets a great attack. �a5 20.'I1;!ie2 dxe5 2 1 . iLxf7t! �xf7 22.'I1;!ih5t �e7 23. lLl e4! After bringing the knight into the attack Black is defenceless. iLd7 24.�h4 exf4 25 .�xf6t �e8 26. l.'!g7 1-O Hamdouchi - Bologan, Belfort 2002. 12.0-0

12 .h4? ! . Stopping Black's h-pawn so drastically is not a recommendable idea. 12 . . . lLl c6 1 3 .0-0 iLd7 14 .lLldl iLe7 1 5 . �h2 �b4 1 6.�e3 iLd8! . A typical manoeuvre to bring back Black's queen. 17 .�d3 lLl a5 1 8 .a3 �b5 19 .�d2 lLlxb3 and Black had a large advantage in Luther - Sutovsky, New York 1 998.

12 ... �d7 13.�hl A useful practical move. 1 3 .5 is more

direct. 1 3 . . . iLe7 ( 1 3 . . . lLl c5 14 . �hl iLd7?! . I think this i s the reason for Black's problems. It is better to try Be7 here. 1 5 . l.'!ab l b5 . Black's king should not be safe in the centre, and Kotronias shows a way to prove it. 1 6.fxe6 fxe6 17.e5 !± with an attack in Kotronias - Hincic, Yerevan 2000.) 14 .'I1;!id4 b5 1 5 .l.'!f3 iLb7 1 6.fxe6 fxe6 1 7. l.'!h3 h4 1 8 .iLg4 �f7 1 9. 1Ll d5 White had a big attack in Guseinov -Villavicencio Martinez, Stockholm 2002 13 . .. h4!

It is very useful for Black to advance his h-pawn in this early stage of the game, because otherwise later White might have had the opportunity to place his rook on h3. 13 .. . b6. A dubious move. Now the black queen is getting in trouble. 14 . � d4! lLlc5 What else? 1 5 .�xf6 l.'!h7 1 6.5 iLe7 17.�d4 iLb7 1 8.fxe6 fxe6 19 . l.'!ab l !

40 The Najdorf

with a strong attack for White in Sulskis - Sutovsky, Koszalin 1 995. 14.h3

Forced. After 14 .!g4 h3! 1 5 .!xh3 �xh3! ' Yes of course! With an open king White will have a lot of trouble in the future. 1 6.gxh3 b5� and Black is doing very well. 14 ... .ie7!

14 . . . b6 1 5 . �adl ( 1 5 .\1!Yd4 was much better here. The lines are similar to the above-mentioned game Sulskis-Sutovsky.) 1 5 . . . !b7 16.f5 �cS I7.fxe6 fxe6 1 S .�d4 ( 1 S . � e3 \1!Yb4! causes White trouble) I S . . . !e7 19 .!g4 lilc5 20. �8 lilxb3 2 1 .cxb3 �c5 22.!xe6 �xd4 23. �xd4 �c5 24. lil d5 and White was better in Luther -Kasimdzhanov, Essen 2002. 15.gadl b6 16.We3

It is difficult to decide where to put White's queen. In this position the e3 square seems to be the right one. 1 6.�d4!? lil c5 ( 1 6 . . . ib7? 17. lil b l ! \1!Yxal I S . lil ld2!±) 17.e5 dxe5 I S .fxe5 f5 with an unclear game. Anyway, Black is a pawn up so White has to prove something. 16 . . ..ib7 17.£5

17. �d4? �cS and the threat of \1!Yb2 was very unpleasant for White in the game Wells - Zhang Zhong, Szeged 1 997. 17 .. J�c8 18.fxe6 fxe6 19..ig4!

Finally White has managed to target Black's weakest spot: the e6-pawn! 19 .. . Wb2

change the outcome of the variation. It just shows once again how strong White's attack is once he has access to the black king. 20Ji:d3 f5 2 1 .exf5 lil e5 22.f6 lilxg4 23. \1!Yxe6 lilxf6 24. �e3! . Bringing the rook to the e-line causes great problems for the coordination between Black's pieces. 24 . . . �h7 (24 . . . lil gS 25 . lil d5 wins instantly for White) 25 . \1!Yxf6 @dS 26. �xe7 �xe7 27. �el d5 2S .a3! Now Black loses material. 2S . . . �xa3 29. lilb l �b4 30.c3 and White i s winning.

1 9 . . . @f7? 20. lild5 ± does not look great for Black either.

1 9 . . . lilc5 . This does not help Black either, as any computer can show. 20.�8 f5 2 1 .exf5 ix8 22.gx8 �c6 23.f6! !xf6 24.!xe6 and White has a clear advantage. 20.EM3!

The standard reply. 20 ••. £5!

19 . . . \1!Yb4. This move does not greatly \i!!iJb 21 .0!::!0 H


Here Nigel Short went wrong. But back in 1 995 chess was a different game than it is nowadays. Today we can all check these lines with our computers and find out how White is winning in this position. 2 1 .exf5L This is the way! 2 1 . . .tLle5 22.f6! 'i;Yxc2 (22 . . . tLlxg4 23. 'i;Yxe6 tLlxf6 24J !e3 is very similar to an above mentioned line. White also has a very strong attack here) 23.f7tL This unpleasant check disturbs all Black's coordination. 23 . . . tLlxf7 (23 . . . @ffi 24.i.e2 tLlxd3 25 . tLld4+- and 23 . . . @d7 24.i.xe6t @c7 25. tLld5t i.xd5 26Jk3t+-) 24 J U2. The point. Black's queen is trapped. 24 . . . VMxd3 25 .VMxd3 tLl e5 26.'i;Yd4 tLlxg4 27.hxg4 @d7 28 J��f7 e5 29.VMxb6 �b8 30. tLlc5t dxc5 3 1 . 'i;Yxc5 �he8 32. tLld5 1-0 Luther - Quezada, Merida 2003. 21...YlYxblt

2 1 . . . 'i;Yxc2? 22.i.d1 f4 23. 'i;Yf3 tLle5 24.i.xc2 tLlxf3 25.gxf3 i.f6 26. tLla5 ! bxa5 27. �xb7 �xc3 28.i.a4t @ffi 29. �xd6± would have given White a very pleasant advantage. 22.�xbl fxg4 23.hxg4 h3?!

Probably this spoils Black's advantage. Other possibilities were 23 . . . �xc2 and 23 . . . tLl e5 with complicated play. However all my computer programs favour Black. 24Jk3?!

24.gxh3! was a good try to win the game. 24 . . . tLl e5 (24 . . . �xc2 25 . �c3! �xa2 26. �c7± i.xe4t 27.'i;Yxe4 �xh3t 28. @gl �xb3 29. tLld2 �xd2 30. �c8t @f7 3 1 . VMh7t @f6 32.'i;Yh6t @e5

33. 'i;Yxd2 and White has wmnmg chances.) 25 . �c3 tLlxg4 26.'i;Yxb6 �xc3 27. tLlxc3 �xh3t 28. @g2 �xc3 29.VMxb7 and White is better here. 24 ... hxg2t 25.cj{xg2 gg8 26J�xc8t hcs 27.cj{fl gxg4 28.� 1d2 e5! 29.YlYc3! i.b7 30.YlYc7

And in this still very complicated position the players agreed a draw. ¥2-lh

42 The Dragon

The Dragon

- By Mikhail Golubev

Shortly before this chapter was completed, an almost anonymous email appeared in my mailbox. The sender proclaimed that The Dragon is

refuted, and attempted to prove it. As this issue seems to be of interest to many, I decided to discuss both the Dragon and its refutation in this introduction.

The Sicilian Dragon is defined by the sequence 1 .e4 c5 2.ttlS d6 3.d4 cxd4 4.ttlxd4 ttlf6 5.ttlc3 g6. It is virtually the only system I play with both colours (and inevitably more often with Black) . Being more or less unable to propose a complete antidote to any of my pet systems. I consider the Sicilian Dragon to be an exception. This opening is one of the few whose theoretical side attracts me more (meaning: to search for the

best moves for both colours) than its practical side. It is hard to explain why. Perhaps it is because the basic conflict is outlined very nicely. With 5 . . . g6 Black builds up an inferior pawn structure with the hopelessly weakened d5-square. (Certainly the Dragon structure is less reliable for Black than that of the Scheveningen. Once, many years ago at a juniors training session, I tried to discuss this with Boris Gelfand, but only half seriously) . Simultaneously, however, Black also develops his kings ide bishop as aggressively as possible. If White does not attempt to exploit the disadvantages of his opponent's set-up in the most principled way, then the activity of the dark bishop can easily turn into a more important factor than the pawn structure.

As practice has showed, White's only critical reply to the Dragon starts with the moves 6.i.e3 ig7 7.S, followed by 8. Vff d2 and then queenside castling. This paradoxically places the white king on the more dangerous side of the board: in the sphere of influence of both the Dragon bishop and Black's queenside rook, which is destined to emerge on the semi-open c-file sooner or later. Paradoxical, yes, but there is simply no other way for White to meet Black's initial strategic threats. 7.f3, which both prevents . . . ttl g4 and removes the pressure on the e4-pawn, would leave White without a clear plan if played together with kingside castling.


Black as a rule answers with 7 • • • 0-0 8.YlYd2 �c6, or 7 • • • �c6 8.YlYd2 0-0, which is of course the same thing (the only really important sideline is 7 . . . lLlc6 8JMid2 id7) . The most common approach is now to try to use all possible tactical tricks to open lines on the queenside and bring displeasure to the white monarch. At White's disposal we have the logical schematic attack on the kingside with h4, g4, h5, etc. As in every opening where both sides have clear aims and targets, the Dragon is strategically simple, but tactically very complex. What is strictly defined. Only How is a real question. Under such circumstances, ambitious amateurs can have a real chance to beat lazy professionals, which they use from time to time!

The current state of affairs is that Black experiences difficulties after both of White's main moves: 9 .0-0-0 and 9 .ic4 (stopping 9 . . . d5 entirely) .

To choose 9.0-0-0 as the main recommendation was not difficult. Firstly, I consider it to be at least no weaker than 9 .ic4. And secondly, 9 .0-0-0 is slightly easier to prepare and play, as the amount of accumulated material and the number of sensible answers for Black is somewhat lower.

The two sharp lines 9 • • • id7?! and 9 . . . �xd4 10..bd4 .te6 both see Black attacking in the usual Dragon spirit. But both are in fact dubious, especially the first, while the latter allows White to use

positional means to fight for an advantage if he wishes.

9 • • • d5! is, in my view, Black's best move. However, this allows White to exploit his pluses strategically. Extraordinary attention to the safety of his king is usually not required. White has several popular options here (Dvoirys' move l O.'�el ! ? is still interesting) but I will concentrate on the main line 10.ad5 �xd5 1 1 .�xc6 bxc6 12.J.d4!.

It is true that lately the claims of the "refutation" of the Dragon were more often related to White's particular findings in the sharpest sub-lines of the 9. ic4-system. However, a simple example (which should remove unrealistic expectations of this chapter) is the line 9 .0-0-0 lLlxd4 1O .,txd4 VNa5 1 1 .,tc4! ? 1 1 . . . ie6 12 .ib3, which is a part of the repertoire. This old and now rare deviation from the main lines is also an integral part of the 9. ic4 system, where it arises after 9 . . . lLlxd4 10 .ixd4 ie6 1 1 .,tb3, represents approximately 3% of the mass of material related to 9. ic4.

It makes no practical sense to speak about a guaranteed win in this somewhat better strategic position with its almost never-ending possible continuations. Very little in chess can be fully proved or refuted. All theory is based on evaluation, comparison, probability and similar uncertainties. The task for a serious player preparing a specific line for White is to find reasons and variations that

44 The Dragon

will convince himself that he will be able to get an advantage. Only someone located above will really know if it will be sufficient to win or not.

With Black we usually aim at convincing ourselves that we can achieve clear equality with our openings. And if we do not succeed, we switch to other lines that we hope are better. I believe it was this and not a refutation, which caused such giants as Alexander Khalifman and Kiril Georgiev to abandon the Dragon. One way or another, I limited the aim of my work to prove at least a small advantage for White in the most critical lines. And this can still only be achieved as a wish, as

no writings on the opening can ever claim to be free from vulnerable assessments.

The accuracy of the data here obviously has its natural limitations. I have sometimes changed the original move order of the specified games. And for accuracy on the origins of the novelties, I refer the readers to their electronic databases. The recent CD by Dorian Rogozenko B75-B79 is, by the way, absolutely wonderful. It is the first commercial Dragon database that I not only merged with my own, but also tend to use separately.

Finally a few words of thanks: I am grateful to Emil Wellner for sending me his interesting analyses, as well as to Ilia Balinov, Erik van den Doel, Jonathan Rowson, and Bogdan Lalic who agreed to answer questions regarding their games. My friend, International Master

Carmen Voicu, helped me significantly in analysing a couple of tricky variations, and for that I am grateful.

Now to the games.

Game 12 Palac - LaUe Pula 2000

1 .e4 e5 2.tlJa d6 3.d4 cxd4 4.tlJxd4 tlJf6 5.tlJc3 g6 6 • .ie3

6 . . . 'Llg4?? now loses to 7.ib5t! , so there is no reason to begin with 6.f3. 6 •.• .ig7

6 . . . 'Ll c6 has virtually no independent significance. 7.f3, and now 7 . . . ig7 is the most normal, and 7 . . . id7 will transpose to the early . . . !d7 after 8 .'�! d2 Ei:c8 9.0-0-0 ig7, a main sideline for Black in the Rauzer Attack, which will be considered below in the note to Black's 8th move.

An unusual move order from Kasparov - Georgiev, Sarajevo 2000 is related with


another topic - the early . . . a6. After 5 . . . a6 6.i.e3 ttlbd7 7.f3 g6 (which is equivalent to 5 . . . g6 6.i4e3 ttl bd7 7.f3 a6) 8 .\Wd2 b5?! (8 . . . i.g7 transposes to the 6 . . . i.g7 7.f3 a6 lines) , White of course played 9.a4! bxa4 (9 . . . b4 1 0. ttl d5!) 1 O . ttl xa4!? i.g7. Later Kasparov proposed l 1 .c4! with advantage, as the most precise. 7.a!

Rauzer's concept, which time has proved to be the best. White controls both e4 and g4 and the f6-knight now becomes a passive, defensive piece. At the same time f3 builds the basis for a future attack with g2-g4 and h2-h4-h5. 7 ... �c6

7 . . . 0-0 makes no difference if Black wishes to play the main lines with 8. \W d2 ttl c6. An independent line for Black after 7 . . . 0-0 8 .\Wd2 is 8 . . . d5?! 9.e5 ttle8 (9 . . . ttl fd7 1 0.f4 should give White some advantage after all reasonable moves, e.g. 10 . . . ttl b6 1 1 .i4e2!? ttl c6 12.0-0-0, etc.)

1 0 .f4 f6. Now 1 1 .0-0-0 fxe5 12 .fxe5 ttl c6 1 3 . ttl f3 i.g4 14. ttlxd5 is a very common recommendation, but 14 . . . e6!it gives Black chances to survive in a slightly worse endgame. So I would rather advise White to deviate with l 1 . ttl f3, and if 1 1 . . . fxe5 then 12. ttlxe5! as in Gufeld-Zimin, USSR 1 958 .

More grounded is another sideline with 7 . . . a6 (a hybrid between the Dragon and the Najdorf) . The disadvantage of the . . . a6, . . . b5, . . . b4 plan in the Dragon is evident: it does not really help Black to gain control over the d5-square. Still 7 . . . a6 was tested by some of the greatest chess players ever. But, starting from the 70s, the line began to be forgotten. Yet recently it has begun to reappear occasionally at grandmaster level. Here I recommend 8 .\Wd2 (the immediate 8 .i.c4!? is a decent alternative) 8 . . . ttl bd7!. The early . . . a6 can only be justified with play in the spirit of the Najdor£ (The hasty 8 . . . b5 is considered to be insufficient in view of 9.a4 b4 10 . ttl a2 a5 1 1 .i4b5t i4d7 12 .c3 bxc3 1 3 .ttlxc3 .

After 8 . . . 0-0?! 9.0-0-0! followed by h4 White has great chances of developing a crushing attack on the kingside. He scores over 80% from this position! An illustrative line is 9 . . . b5 1 O .h4 h5 l 1 .g4! e5?! 12. ttl b3 hxg4 1 3 .i4g5 , etc.

h is important to note that 8 . . . ttlc6?! in conjunction with . . . a6 is always dubious, and rather devalues the . . . b5 idea. In the normal . . . ttl c6 lines Black plays . . . a6

46 The Dragon

only somewhere around move 14, which is usually a sign that he has run out of constructive ideas in the position.) 9. ic4!? Several old sources, for example Geller in ECO in 1 984, gave a clear preference to this move, while in other lines the bishop remains passive. And now:

a) 9 . . . hS prevents 1O .ih6, but White's position becomes pleasant. Amongst other ideas he can consider castling kingside.

b) After 9 . . . lLl cS 10 .ih6!? might be recommended.

c) 9 . . . V!!c7 10 .ib3! bS (or 10 . . . 0-0 1 1 .h4!? with prospects for an attack, Bilek - Simagin, Budapest 1 96 1 , and 10 . . . h6 1 1 .0-0-0 lLlb6? 1 2.eS!± Karjakin - Romero Holmes, Pamplona 2003 with the idea 12 . . . dxeS 1 3 .lLldbS axbS 14 . lLlxbS V!!d7 l S .V!!b4) 1 1 .0-0-0 ib7 12 .ih6 ixh6 (or 1 2 . . . 0-0 13 .h4!) 13 . V!!xh6 with initiative, Torre - Fuller, Australia 1 975.

d) 9 . . . bS 10 .ib3 ib7

This move transposes to a position which often arises via 8 .ic4. (On 1 0 . . . lLlcS then l 1 .idS!? is funny, but 1 1 .ih6! seems to be more solid. 10 . . . hS ? ! 1 1 .a4 i s clearly better for White, as is probably also 10 . . . 0-0?! 1 1 .a4! with the ideas 1 1 . . .b4 12 . lLl dS and 1 1 . . .bxa4 1 2. lLl c6 V!!e8 13 .ixa4 ib7 14 .eS! .) 1 1 .ih6!? (evaluated as strongest by Botvinnik, who also faced 1 1 .0-0-0 over the board) 1 1 . . .0-0. "Dangerous" according to Botvinnik. (Krutikhin -Botvinnik, Moscow Spartakiad 1 963 continued 1 1 . . .ixh6 12 .V!!xh6 lLlcS 1 3 .0-0-0 lLlxb3t 14.cxb3!? V!!b6 l S . @b1 0-0-0 1 6.b4! and White created some pressure. After this experience of struggling for half a point against his not too famous opponent, the 6th World Champion simply stopped playing the Dragon) 12 .ixg7 @xg7. Approximate, overoptimistic analysis may continue 1 3 .0-0-0 b4 ( 1 3 . . . lLl cS 14.eS!? dxeS l S . lLl fSt @h8 1 6.V!!h6 gxf5 17Jhd8 Ei:axd8 1 8 .V!!e3) 14. lLl a4 (if 14 . lLldS lLlxdS l S .ixdS ixdS 1 6.exdS , then Black has 1 6 . . . V!!aS!) 14 . . . Ei:b8 l S .h4! ( 1 S .V!!xb4 ixe4°o, followed by 16 . . . ia8 and 17 . . . eS) l S . . . eS? ! 1 6. lLl e2 V!!aS 17.hS! ic6 1 8 .hxg6 fxg6 19 . V!!xd6 ixa4 20.V!!e7t @h8 2 1 . ixa4 Ei:fe8 22.V!!f7 Ei:f8 23.V!!xh7t! lLlxh7 24. Ei:xd7+-.

The conclusion is that in the lines with 7 . . . a6 Black has little influence in the centre, and is therefore somewhat worse. ih6 is, as a rule, an important resource


for White. When Black casdes kingside White usually should begin his kingside attack with h4! rather than g4. 8. ti' d2 0--0

If Black likes the idea of 8 . . . 0-0 9.0-0-0 !d7, then 8 . . . !d7 9.0-0-0 [k8!? looks a significandy more flexible move order. The point is that in reply to g4 Black gets the additional possibility of an early . . . h5!? (GM Sergei Tiviakov's speciality) , urging White to close the kingside by g5 (h3? will lose a pawn after . . . hxg4) . Still, after g5 Black will be forced to retreat a knight to h7 (e8 is still occupied by the king!) where it will become extremely passive. White can then develop an initiative with f.3-f4-f5!. Let us consider all of this with more details.

I suggest 10.g4 which is more popular than other moves. Now 10 . . . 0-0 of course transposes to 8 . . . 0-0. Instead Black may play 1O . . . h5 at once, or wait for h4 with 1O . . . lLle5. After the first of these White

scores extremely well, while the latter seems to be more critical. This is because in positions where the kingside is closed . . . lLl e5 can be more useful for Black than h4 for White. Thus:

a} 1 O . . . h5 1 1 .g5 lLlh7 12.f4 (after 12 . cJ1b1 0-0, then Xie Jun's suggestion 13 .lLlb3!? allows quite a typical reply for these lines 1 3 . . . !xc3!? If White wishes to delay f4 then 12.!e2 is reasonable.) 12 . . . 0-0 ( 12 . . . lLlf8? 13 .f5. Untested is 12 . . . !g4!?) 13 .!e2 (also of interest is 1 3 .f5!? and 1 3 . . . lLle5 14.!e2 E:e8 shifts to Yagupov - Motylev below. If Black instead plays 13 .. JWa5 14. cJ1b1 gxf5, then both 1 5 .lLlxc6! ? bxc6 16.lLld5 �d8 17.lLlf4 E:b8 1 8 .b3 fxe4 19 .1Llxh5 !g4 20.,te2 !xh5 2 1 .!xh5 Coleman, and 1 5 . lLlxf5!? !xf5 16.exf5 �xf5 17.!d3 �a5 18 .lLld5!? �xd2 19 . E:xd2 e6 20. lLlf4 h4 2 1 .g6 with initiative for the pawn, Coleman - De Holanda, corr. 2000 are promising for White. Imprecise is 13 .cJ1b1 ,tg4! 14.,te2 lLlxd4 1 5 .!xd4 !xd4 16.�xd4 e5! - Xie Jun) 13 . . . E:e8 (or 13 . . . �a5 14. cJ1b1 lLlxd4 1 5 .!xd4 !xd4 16.�xd4 ,te6 17.f5! Jandek - Binas, corr. 1 998-2000) and now White should probably play 14.f5!? (why not?) 14 . . . lLle5 (or 14 . . . �a5!? 1 5 .cJ1b1 and then for example 1 5 . . . ,txd4 16.,txd4 lLlxd4 17.�xd4 �c5 1 8 .�d2 �e5 1 9. E:hfl ;!; Reichardt - Berclaz, corr. 1 998-9) 1 5 . E:hfl with pressure after 15 .. . cJ1h8 16 .lLld5 Yagupov - Motylev, Russian Ch 1 998, or 1 5 . . . a6 16. E:f2!? b5 17 .E:fd1 - Coleman.

48 The Dragon

b) 1 0 . . . ltle5 1 1 .h4 h5!? (after the rare 1 1 . . .b5 the principled move is in many ways 1 2. ltl cxb5. Then: 12 . . . 0-0 13 .h5 ltlxB 14. ltlxf3 ixg4 Mestel - Christiansen, Hastings 1 978/9, and now 1 5 .ie2 ltlxe4 16 .�e 1 ! Nunn.) 1 2.g5 ltl h7 1 3 .mb1 !? 0-0 ( 1 3 ... ltl f8 14.f4 or 14 .ie2!? ltl e6 1 5 . ltlxe6 i.xe6 1 6.id4 0-0 1 7. ltld5 ixd5 1 8 .exd5 ltl c4 19 . �b4! Istratescu - Gelashvili, Kallithea 2002) 14 .i.e2 ltl c4 1 5 .ixc4 l:hc4 16 .�d3 �c8 (Zuidema - Bilek, Havana (Ol) 1 966) 17.f4! ?t Matulovic.

Before going on with 8 . . . 0-0 it is useful to observe a recent trend: in a number of games Black tried to combine an early . . . .td7 with an early . . . h5, without even waiting for White's g4 or h4 (e.g. 8 . . . h5 , or 8 . . . .td7 and 9 . . . h5) . I always used to think that such ideas are unsound for Black, because White will quickly play h3!? (and develop the fl -bishop, if necessary) , intending g4. If Black then allows g4 the pawn will create colossal strategic pressure on the f6-knight, which Black can hardly survive. So, after h3 Black's only logical continuation will be . . . h4 (forgetting completely all ideas involving castling, as then the h4 pawn will be doomed) , followed by moves like . . . ltlh5 or . . . �h5.

The move order nuance is that 8 . . . i.d7 9.0-0-0 h5 allows White to develop his bishop to c4 (which is promising here and not really transposing to the 8 . . . 0-0 9 . .tc4 Iabyrinth) , while in the line which

we just considered above, i.e. 8 . . . id7 9.0-0-0 �c8, the main move 1 O.g4 prevents Black from the . . . h7 -h5-h4 advance, while 10 . mbl allows it. Black's entire concept looks so strategically unsound to me that I will refrain from further details. Normally, if Black plays with the king in the centre, he should collapse quite quickly. 9.0-0-0

"Strong and safe", comments GM Boris Alterman, one of the greatest experts in the Dragon, who amazingly has a negative score against White's queenside castling. Alternative 9 . .tc4 prevents 9 . . . d5 completely, but gives Black some time to prepare his actions, and is therefore much more complicated.

9 .. .J.d7?! "Fundamentally unsound" - FM Stefan

Sieveres, "a flagrant error" - 1M Attila Schneider.

Now I partly agree with these strong


statements. The second player takes great risks with this move, while White faces no pressure and can calmly start his kingside assault, as no adequate counter-plan for Black can be seen. Even the common but rather innocent idea from the 9 . .ic4 line: . . . Jk8, . . . lLl e5 and . . . lLl c4 looks senseless here, as Black will lose two tempi in comparison with the sharp positions after 9 . .ic4.

As usual in the Dragon, the advance . . . b7-b5-b4 sends a rather pleasant invitation to the white knight to visit d5 . From another point of view, 9 . . . .id7 is in some ways the most complex of the available moves for a very simple reason: Black does not exchange pieces! This factor alone cannot change the assessment of the line, but (with the exception of the very top level of chess) it significantly improves the practical chances for an ambitious Black player. In the last decade theory has begun to develop intensively in this line. White has started to learn how to avoid the opponent's tricks, and Black's choice of playable looking lines has gradually narrowed.

The system with 9 . . . .id7 is highly transpositional, so here the first player should definitely know the evaluation of at least 3-4 key positions (both promising and unpromising) . Such knowledge will be essential for navigating during a practical game. Before we dive into variations, I should also note that the value of the developing move .ie2 is

often underestimated. In fact e2 is a very good square for the bishop! Thus, the continuation of Svidler-Golubev, BadenBaden 2002: 9 . .ic4 .id7 10.0-0-0 lLl e5 1 1 ..ie2! ? , where a top 10 regular puts his bishop on e2 even with the loss of a tempo, should at least make us think. IO.g4!?

We will concentrate on this, the most popular continuation. IO •• J:�c8

1 O . . . lLl e5 1 1 .h4 as a rule transposes after 1 1 . .Jk8 . The most common option to do without . . . Ek8 is 1 1 . . .b5 ( 1 1 . . .�a5 can be answered by lLl b3 at some point, or by 12 . 'it>b1 �fc8 13 . .ie2!, Grischuk - Soloviov, Russian Ch 1 999, which is akin to Macieja's important game below.

On 1 3 . . . b5 Grischuk recommended 14. lLlcxb5 \&d8 1 5 .h5 Too risky, perhaps, is 1 1 . . .h5 1 2 . .ie2!? or 1 2.gxh5 lLlxh5 1 3 .�gl ) . After 1 1 . . .b5, the line goes on with 12.h5 ( 1 2. lLl d5 is also good.)

50 The Dragon

b4 ( 12 . . . lLlxf3 13 .lLlxf3 .txg4 is hardly sound after 14 . .te2!?) 1 3 .lLld5 lLlxd5 ( 1 3 . . . e6 14 . lLlxf6t 'i4'xf6± and now maybe 1 5 . .te2 - but not 1 5 . .tg5? lLl d3t!) 14.exd5 'i4'a5 1 5 . �bl 'lM!xd5 1 6.hxg6 ( l 6 .lLl f5!? is another dangerous move for Black which GM Vladimir Chuchelov and I analysed some 10 years ago. Even now I am not quite sure about 1 6 . . . '1M! xd2 17 . lLlxe7t �h8 1 8 . .txd2 l:!fe8 19 .h6 .tf8!? 20. lLld5 lLlx£3 2 1 ..txb4 .txg4 with the idea 22. lLl c7 l:!e4!) 1 6 . . . fxg6 17.'i4'h2 with a very strong White initiative: 17 . . . h6 (both 17 . . . h5 and 17 . . . �f7 can be answered in the same way) 1 8 . .te2 !±

Bologan-Fedorov, Elista (01) 1 998. 1 0 . . . 'i4' a5 and now l 1 . lLl b3! ? '1M! c7 1 2.g5 lLlh5 13 .ie2!? l:!ac8 14. lLld5 'i4'd8 1 5 .f4 e6 1 6. lLl c3 lLlb4 17.�b l ! Nevostrujev Soloviov, Samara 2000. Instead 1 1 ..tc4! ? transposes to the 9 . .tc4 'i4'a5 system. 1 l .h4

A very common move. Also interesting is: 1 1 . �b l ! ?

( l 1 ..te2!? lLl e5 12 . �bl jusuransposes) lLl e5 12 . .te2. This is the pet line of GM Oleg Korneev, which has been used by him at least four times. Virtually Black's only sensible reply is 1 2 . . . b5! ? (discouraging is 1 2 . . . h5? ! 13 .h3! or 1 3 .gxh5 lLlxh5 14.f4 lLl c4 1 5 . .txc4 l:!xc4 1 6.f5 Korneev -Belezky, Lorca 200 1 , or 1 2 . . . a6 13 .h4 h5 14.gxh5!? lLlxh5 1 5 . l:!hgl with initiative. And after 12 . . . 'IM!a5 , then 13 .h4 is good, and even better is 1 3 .lLlb3! 'i4'c7 14.g5 lLlh5 1 5 . lLl d5 'lM!d8 16 . .txa7 Korneev -Getta, San Sebastian 2000) . Now White can consider:

a) 1 3 .lLldxb5 was tested in the stem game Korneev - Fedorov, Krasnodar 1 998 : 1 3 . . . ixb5 (worse i s 1 3 . . . lLlc4?! 14 . .txc4 l:!xc4 and now maybe 1 5 .e5!?) 14 . lLlxb5 l:!b8 (l4 . . . a6! ? Fedorov) 1 5 .c4 (brave is 1 5 . .txa7!?) 1 5 . . . a6 16. lLl d4 'lM!c7 17. l:!c 1 . Here Black could prevent the c4-c5 advance with 17 . . . lLl fd7! and if 1 8 . lLl b3, then 18 . . . lLlc5! with compensation -Fedorov.

b) 13 . lLl cxb5! ? was tried by Korneev recently. It seems that 1 3 . . . .txb5 14 . .txb5!? is the idea. Instead, Korneev - Carlsen, Reykjavik 2004, continued with 13 . . . a6 14. lLlc3 lLl c4 1 5 . .txc4 l:!xc4 when Black had no real compensation. 1 6. lLl de2! ?± could be a move here.

c) 1 3.h4!? makes sense as well. 1 3 . . . b4 (on 1 3 . . . lLl c4?! follows 14 . .txc4 bxc4 1 5 .h5 ± . 1 3 . . . 'IM!a5! ? is considered in Macieja's line below) 14 . lLld5 lLlxd5 1 5 .exd5 'lM!a5 ( l 5 . . . lLlc4?! 16 . .txc4 l:!xc4


17.h5 ±), and now untested is 1 6.h5! with the idea 16 . . . V�!fxd5 17. CLl fS! . This is probably stronger than 1 6. CLl b3 Vlic7 17.h5, which transposes to the variation with 1 5 . . . b4, deviating from the main game.

1 1 . • • �e5 The main alternative is 1 1 . . .h5. Now

12 . CLld5!? is perhaps the most dangerous move for Black. I remember analysing it for a really long time in the army, fourteen years ago. Since then no really important games have been played with it. The most significant of the lines, already published by me in Correspondence Chess Informant and New In Chess, are:

a) 1 2 . . . hxg4 1 3 .h5 (much stronger might be 1 3 .CLlxf6t!? i.xf6 14. CLlxc6!, or 1 3 .CLlxc6! ? followed by 14 . CLlxf6t i.xf6 1 5 .h5 after any Black recapture) 1 3 . . . CLlxd5 ( I 3 . . . CLlxh5!? and now after both 14. CLl fS Schneider & Sapi and 14 . CLlxc6 bxc6 1 5 . CLl f4!? not everyone

would be convinced about White's chances) 14 .exd5 CLlxd4 1 5 .i.xd4 Hazai - Bilek, Budapest 1 98 1 . Now 1 5 . . . i.xd4!? 1 6. V�!fxd4 Wlb6 and Black is hardly much worse. However those 1 3th move alternatives need to be investigated in practice.

b) 1 2 . . . CLlxd5 1 3 .exd5 CLlxd4 14 .i.xd4 i.xd4 1 5 .Vlixd4 and now 1 5 . . . Wla5 1 6.gxh5 Wlxa2 17.h6 f6 18 .i.d3± or 1 5 . . . hxg4 16.fxg4! .

c) 1 2 . . . CLlxd4 13 .i.xd4 e5 (I3 . . . hxg4 14 .i.xf6 i.xf6 1 5 .h5 g5 1 6.f4!? , less clear is 16.e5 dxe5 17.CLlxf6t exf6 1 8 .Vlixd7 Wlxd7 19 . E1xd7 gxf3 and the black pawns are at least frightening) 14 .i.e3 CLlxd5 (Shianovsky-Geller, Kiev 1957) 1 5 .Wlxd5 ! hxg4 1 6.h5 and I evaluated this in White's favour.

After I 1 . . .Vlia5?! then 12 . CLlb3!? looks unpleasant for Black. 12.h5 Again, the dominating choice in practice. And again, not necessarily the best.

52 The Dragon

The position after 12 . i>b1 is known since Boleslavsky-Geller, Zurich Ct 1 953. I suggest White to avoid it : 12 . . . h5! seems to be playable for Black.

Better is 12 .i.e2!? and now: a) 12 . . . h5 is very risky here: 13 J �dgl ! ?

(recommended by Schneider & Sapi) 1 3 . . . ltJc4 ( 1 3 . . . b5 14.gxh5 b4 1 5 . ltJ d5 ±) 14 .i.xc4 l:'l:xc4 and now, probably, 1 5 . ltJ f5 ! ? with an attack.

b) 1 2 . . . b5!? 1 3 .i>b1 ( 1 3.h5 VNa5 is the Smeets - Carlsen game, see 12 .h5 VNa5 1 3 .i.e2) and we are back in Korneev's line, which after 13 . . . � a5 in its turn transposes to 12 . . . VNa5 below.

c) 1 2 . . . �a5 1 3 .i>b l !

This i s one of these important positions, which can be found at the crossroads between a lot of lines. It first appeared in Macieja - Grabarczuk, Sopot 1 997. Black faces problems here: 13 . . . b5 is critical, but it gives White a maximum of possibilities : He can grab the b5-

pawn, or choose between three other lines represented in the games Smeets -Carlsen, Lau - Tiviakov and Palac - Lalic. ( 1 3 . . . l:'l:xc3 14.VNxc3 VNxc3 1 5 .bxc3 l:'l:c8 is a typical Dragon endgame, where Black's compensation for the exchange is insufficient. To feel really confident after the . . . l:'l:xc3 sacrifice, Black should ensure himself of either an extra pawn on the kingside, or the preservation of the queens on the board. Now 16 . i>b2!± was Sutovsky - Hodgson, Oxford 1 998.) 14 . ltJ b3!? ( 14.h5!? transposes to Smeets -Carlsen - see this game below) 14 . . . VNc7. Here a line is 1 5 .g5 ( 1 5 .h5!? transposes directly to the main game) 1 5 . . . ltJh5 ( 1 5 . . . b4? loses a piece after 1 6. ltJ b5 i.xb5 17.gxf6+- Lau - Tiviakov, Montecatini Terme 1 994) 16 . ltJ d5 �d8, followed by the principled 17.i.xa7!? ltJc4 1 8 .i.xc4 bxc4 1 9 .i.b6 �e8 20. ltJ d4. White has an extra pawn, but unfortunately I am afraid to recommend such a greedy choice to White. The position is complex and some global investigation is required. 12 ... Wa5!

1 2 . . . ltJx£3 13 .ltJxf3 i.xg4 is considered to be insufficient: 14 .i.e2 (I like this more than 14.h6) 14 . . . i.xh5 and now the simplest is probably 1 5 . ltJ d4 i.xe2 1 6. ltJ dxe2! ( 1 6. VNxe2 l:'l:xc3 17.bxc3 ltJxe4 1 8 .i.h6 Sermek - Kovacevic, Belgrade 1 989 allows 1 8 . . . ltJxc3!) 16 . . . h5 (or 1 6 . . . ltJg4 17.i.d4 ltJ e5 18 .�e3±) 17.i.xa7!? Hanison - Betts, corr. 2000. 13.tilb3!?


1 3 .�b l ?! (which is the most played move here) is basically what Black hopes for in the entire 9 . . . i.d7-system.

Now comes 1 3 .. .lihf3! (according to Ward, this is a more accurate introduction than 1 3 . . . �xc3) 14. ttJxf3 �xc3 (or 14 . . . i.xg4 first) 1 5 .�xc3 �xc3 1 6.bxc3 i.xg4 17.i.g2 ttJxe4 with fine compensation for the rook, which is confirmed by a few dozen games, starting with Sandor - Aagaard, Hamburg 1 993.

Fresh is 1 3 .i.e2! ? b5 (13 . . . �xc3 fails to impress. 14 .�xc3 �xa2 and now 1 5 . '\1;\1 a3, or the more ambitious 1 5 .�b3!? �al t 1 6. �d2 '\1;\Ia5t 17 .�c3!) and now 14. �b l ! ? (if 14 . ttJb3 then 14 . . . � c7 transposes to Palac - Lalic, while Black's additional possibility is 14 . . . �a6) 14 . . . �xc3 ( 14 . . . b4 1 5 . ttJ d5 ttJxd5 1 6.exd5 with the idea 16 . . . �xd5 17. ttJ 5!± is something we have already seen in Korneev's line. Or 14 . . . ttJ c4?! 1 5 .i.xc4 bxc4± 1 6. ttJ d5!?) 1 5 .'�xc3

�xc3 ( 1 5 . . . b4± is the big OptimIst s choice) 1 6.bxc3 �c8. This was Smeets - Carlsen, Corus-C 2004. Again this is one of these endgames where Black's compensation for the exchange is rather vague. The game continued 17.hxg6 fxg6 1 8 . �b2 a5 19 .i.h6 i.h8 and here Rogozenko gives 20. �b l !± . 13 . . . Wc7 14.!e2 b5 15.�bl!

The position after 1 5 .�b l i s quite complex, but White seems to have good chances. (Still, we remember the promising early deviations: 1 3 .i.e2, 12 .i.e2 and Korneev's 1 1 . �bl ttJ e5 12 .!e2) . Now 16.hxg6 fxg6 17.g5 is already quite a threat, which would be seen after moves like 15 . . . a6? 15 ... tLlc4!?

Another direction is 1 5 . . . b4 1 6. ttJ d5 ttJxd5 1 7.exd5 and now Black must make a difficult choice:

a) 17 . . . 5 weakens the kingside: 1 8 .hxg6 hxg6 1 9 .i.h6 f4!? 20. ttJd4!? '\1;\Ic5 2 1 . '\1;\Iel

54 The Dragon

with a dangerous initiative, Fressinet -Polzin, Bundesliga 200 1 /2 .

b} 17 . . . a5 1 8 .lLld4! (the tempting 1 8 .ih6?! does not work well: 1 8 . . . ixh6 19 .�xh6 �xc2t 20. �a1 �xe2! ! and now 21 .hxg6 �xd1 t 22Jhd1 fxg6 with excellent compensation, Short - Bu Xiangzhi, Taiyuan 2004, or 2 1 .lLld4 lLlxg4! 22.fxg4 and there is not only 22 . . . �e5 23.hxg6 �g7 24.gxh7t �h8°o Sax - Cebalo, Croatia 2002, but also 22 . . . �e4! - Cebalo) 1 8 . . . lLlc4 ( 1 8 . . . a4 19 .ih6±) 1 9.1xc4 �xc4 20.hxg6 (less clear is 20.ih6 ixd4 2 1 .ixf8!? and now 2 1 . . .if6, or even 2 1 . . .ic3) 20 . . . fxg6 21 . �h2. In Hautot - Goormachtigh, Belgium 2003/4 Black should now have played 2 1 . . .h6!, when I can suggest 22.b3!? �xd5 23. lLl f5! with advantage for White, rather than 22.ixh6 ixd4 23.b3!? �c3 24Jhd4 �xd4 25.ie3 �g7 26.id4 �f6 27.g5 if5!.

c} 17 . . . t'tJc4?! 1 8 .1xc4 �xc4 19 .id4± .

d} 17 . . . �fe8 18 . �h2 as in Tomescu -Piva, Porto San Giorgio 2003 may look reasonable, but here Black has 1 8 . . . lLlc4! (instead of the game's 18 . . . a5?) 19 .ixc4 �xc4 when he is alive, because the h2-square is no longer available for the white queen. e.g. 20.hxg6 (20.�dh1 g5 2 1 .1xg5 a5 ! 20. �f2 ie5!?) 20 . . . fxg6 2 1 . �dh1 (2 U : 1hh1 ie5) 2 1 . . . �f8! 22Jl:xh7 �xf3! and now 23. �xg7t only gives a draw.

But, instead of all this, 1 8.lLld4!? preserving the initiative looks interesting.

16 • .hc4 bxc4 1 6 . . . �xc4?! looks terribly bad for Black:

it can be punished by 17.e5 or 17 .ih6. 17.�d4

In the line 9 .ic4 �a5 1 0.0-0-0 id7 1 1 .ib3 �fc8, when White follows with h4, g4, h5 and Black responds with . . . lLl e5 , b5, lLlc4, a quite similar position often arises, which is difficult for Black. Here he can attack b2 faster, but it hardly improves his chances. 17 ••• YlYb7

17 . . . �b8 would normally transpose. 18.YlYh2?!

1 8 . �a1 ! ? �b8 1 9J :1b 1 is probably a more suitable method to protect b2. After that Black's prospects for counterplay would have been limited.

At the same time, 1 8 .ih6! �b8 1 9 .b3 looks a more concrete decision. Of course, b3 is (hypothetically) vulnerable, but White's attack appears to be faster. 18 •• J�b8 19.ic1 h6!?


Prophylactic against the possible 20.hxg6 and 2 1 . lt'l d5 . White's bishop is busy defending b2, so the h6 pawn is untouchable for the moment. 20.hxg6 fxg6 21 .b3

Attacking h6. 2 1 .�e2!? was safer. 21 ... cxb3 22.a:x:b3 �fc8?!

22 .. . g5 ! with double-edged play. 23J:�d3?!

23.i.xh6! is of course tempting, but it is surprisingly hard to make it work. Still, it seems that White is better after 23 .. Jhc3 24.i.xg7 'it>xg7 25 . �h6t 'it>O 26.g5 It'le8 27J':'ih4! e5 28.�h7t 'it>fB 29 . l'i:dhl ! , which i s inevitably followed by 30.�xg6 exd4 3 1 . l'i:h8t 'it>e7 32. l'i:xe8t. 23 ... h5?

23 . . . e5? is refuted by 24.i.xh6! but 23 . . . g5 ! was called for, even more evidently than one move earlier. 24.gxh5 tllxh5

24 . . . e5 25 .h6! ?± . 25.Yfg2!±

Black's king has become too vulnerable. Though the following was not free from inaccuracies, White got to the enemy monarch in the end. 25 .. .i.eS 26J�xh5 .ixd4 27 .�xd4 gxc3 2S..tb2 S:xc2 29.Yfxc2 gxh5 30.Yfg2t 'it>m 31 .gb4 ib5 32.YfgG e6 33.Yfh6t 'it>e7 34.gd4 i.e2 35.e5! ha 36.Yfg7t 'it>e8 37.YfgSt 'it>e7 3S.Yfg7t 'it>eS 39.YfgSt 'it>e7 4O.exd6t 'it>d7 41 .Yfflt 'it>dS 42.Yfmt 'it>d7 43.Yfe7t 'it>c6 44.gc4t 'it>b5 45.d7 1-0

Game 13 Golubev - Poliantsev Mariupol 1 990

1 .e4 c5 2.tlla d6 3.d4 cxd4 4.tllxd4 tllf6 5.tllc3 gG 6..te3 ig7 7.a 0-0 S.Yfd2 tll c6 9.0-0-0

Black's ideal plan after 9.0-0-0 includes the exchange on d4, . . . Wa5 and . . . �e6, taking control of many squares. Fortunately for White he is able to prevent the execution of this plan in its pure form. Now we will start considering Black's options one after another. 9 ... tllxd4

The immediate 9 . . . W a5 is very rare. Then 10 .i.c4!? requires knowledge of 9. i.c4 to which it transposes, and 10 . 'it> b 1 should suit White if he intends to meet 9 . . . lt'lxd4 10 .i.xd4 Wxa5 in the same way. The most radical reaction, however, is 10 . It'l b3!? �c7 and now 1 1 .h4 or l 1 .g4 i.e6 12 .g5 It'l d7 13 .h4 a5 14 . lt'lb5 �c8 1 5 . 'it>bl with advantage, Smirin - V. Alterman, Haifa 1 993. 10..txd4

By far Black's most popular move here is 10 . . . i.e6. It will be covered separately in the next game. 10 ... Yfa5!?

The main idea behind this queen move (used by Geller and other strong players in the 50s) is, of course, 1 1 . . .i.e6!. For example, after I I .g4 it will transpose to the main line of the 9.g4 line, which is acceptable for Black. 1 1 .h4 also

56 The Dragon

allows 1 1 . .. ie6, but there is a fresh idea: 1 2 .�g5 !? Yemelin - Kalashnikov, St Petersburg 2000 should, at least, be mentioned. Usually, White chooses between two other promising, but very different, options. l l .i.c4

Improving the position of the only relatively passive piece. A reasonable alternative is 1 1 . 'it>bl !? , which can lead to tense play after 1 1 . . .e5 (not 1 1 . . .ie6? 1 2 .lDd5) 12. ie3 ie6. e.g. 1 3 .a3 E:fc8! 14. ie2 E:c6 and after 1 5 . lD b5, 1 5 . . . �a4! was found by Wellner. Instead White can consider 1 5 .g4 E:ac8 16. lD d5. 1l . . . ie6 12.ib3!

An important part of White's concept. In contrast 12 .ixe6 would significantly improve Black's influence in the centre.

In this position Black has tried four main plans: a quick . . . b5-b4 (without exchanging on b3) , 12 . . . ixb3 and 1 3 . . . E:fd8 (trying to prepare . . . e5 and

. . . d5) , 1 2 .. J Hc8 and . . . b5-b4 (usually with the inclusion of . . . ixb3 at some early point) , and finally, the less forcing . . . E:fc8 and . . . E:c6 (which also normally involves . . . ixb3) . None of these plans has been found to be satisfactory by Black players. Before entering into greater detail it is important to note that White should not be afraid to capture on b3 with his cpawn (which in fact will remain his main option even if Black will allow him to play 'it>bl first) . There is no way for Black to exchange major pieces without serious concessions, and middlegame factors are of most importance. The capture cxb3 ensures a safe position for the white king, and Black's pawns majority in the centre, as well as his kingside, becomes a target of White's long-term pressure. 12 .. JUc8

1 2 . . . ixb3 1 3 .cxb3 E:fd8 14 .'it>bl Eld7 i s the oldest of Black's ideas. After 1 5 .g4 Elad8 White has usually tried the prophylactic 1 6. � e2, which works well against 1 6 . . . e5? 17.ie3 d5 1 8 .g5 + - . As 16 . . . � a6! ? is not entirely clear to me, I will follow Boleslavsky's main recommendation: 1 6.h4!? e5 17 .ie3 d5 1 8 .exd5 lDxd5 1 9 .1Dxd5 �xd2 20. Elxd2 Elxd5 2 1 . Elxd5 Elxd5 22. Elc1 ! with a better endgame (Goglidze - Toprover, Tbilisi 1 955) .

After 12 . . . b5 13 . 'it>bl b4 14 . lDd5 ixd5 1 5 .exd5 �b5 White's best, according to Rogozenko, seems to be 16 . �d3! �xd3 (after 1 6 . . . �b7 17 . Elhel a5 1 8 .ia4

gabS 1 9 .94 gfcS 20.if2 White is clearly better - Gufeld) 1 7Jhd3 gfcS! ( 17 . . . a5 ? l S .ia4! Gufeld) l S . ge 1 (Here l S .ia4?! allows lS . . . ttlxd5! 1 9 .ixg7 ttl b6!) l S .. Jk7 19 .c3 a5 20.cxb4 axb4 2 1 . gdd1 with a better endgame in Motylev-Felgaer, Linares 200 1 . 13.bl

13 ... ixb3 Or 1 3 . . . b5 14 J :1he l ! ixb3 ( 14 . . .

b4 1 5 . ttl d5! ixd5 1 6.exd5 gc7 17.a4! with a clearly superior position, Bagirov - Gufeld, Leningrad 1 960 14 . . . gabS 1 5 .e5 !±) 1 5 .cxb3. Now:

a) Black has never tried 1 5 .. J !c6. One of the interesting options for White here is 1 6.ixf6 ixf6 17. ttl d5 'lWxd2 1 S . gxd2 g7 1 9 J :1c 1 .

b) Another waiting move is 1 5 . . JhbS!? Emil Wellner analysed 1 6.e5 dxe5 17. gxe5 e6! l S . ttlxb5 ( 1 S .'lWe2?! gdS! 1 9 . 9xb5 gxb5 20. ttlxb5 e5! Wellner, and now 2 1 . 'IW xe5 ttl eS + with the idea

22.'lWe7 if6! 23.'lWe2 a6! , or 2 1 .ic3 gxd1 t 22. c2 'lWb6 23. xd1 ttl d5!) lS . . . 'lWxd2 1 9. 9xd2 ttld5 20. ttlxa7 gc7 2 Uhd5 exd5 22.ie3;!; with a sharp endgame. White's less obliging options are 1 6.ixf6 and 1 6 .a3!?

c) 15 . . . b4 1 6.,txf6! ( 1 6. ttl d5 ttlxd5 17.ixg7 is not a precise move order as it allows an intermediate 17 . . . ttl e3!) 16 . . . bxc3 17 . ixc3 ixc3 ( 17 . . Jhc3? i s refuted nicely by l S .ge2!! gc5 1 9 .b4+-) l S .bxc3 gxc3 ( 1 S . . . 'lWxc3 19 .'lWxc3 gxc3 20. gcl can lead to the same) 1 9. ge3 gacS (Black is also suffering after 1 9 . . . gc5 20.'lWxa5 gxa5 2 1 . gc3 or 2 1 . gcl ) 20. gxc3 'lWxc3 (or20 . . . gxc3 2 1 . b2 gc5 22. 'lWxa5 gxa5 23. gcl ±) 2 1 .'lWxc3 gxc3 22. gcl gxc l t 23 . xc1 . The arising pawn endgame will be in White's favour - his queenside majority offers prospects of sacrificing a pawn there, distracting Black's king from the kingside, and to win the black pawns then. Black, however, can build up some kind of fortress by keeping his f-pawn on f7 and placing his e-pawn on e6. This will prevent immediate access to the black pawns for the white king. Dutch IMs Karel van der Weide and Jeroen Bosch filled eight pages of New In Chess Magazine 1 995/4 with extensive analysis proving White wins even in this case. They showed that every pawn move on the kingside is extremely sensitive. The authors' main line goes: 23 . . . f8 (Not 23 . . . g5 24 . b2 and the white king reaches a6) 24. c2 (24.b4 eS 25 .b5

5B The Dragon

'i!fd7 26. 'i!fc2, proposed by Shereshevsky & Slutsky, allows Black to escape after 26 . . . d5! - van der Tak.) 24 . . . 'i!feB 25. 'i!fc3 'i!fd7 26. 'i!fc4 'i!fc6 27.h4! (Not 27.b4 g5!) 27 . . . h6 2B .b4 e6 29.a4 'i!fb6 30.f4! (not 30.g4? g5) 30 .. . 'i!fc6 3 1 .g4 'i!fb6 32.5. White will advance his pawns to f6, b5 and a5, then he will push e4-e5, which after . . . d5 will allow him to move the king via c5 and d6 to Black's f7 pawn, and promote his f6 pawn. In his turn, Black will get time to promote his d5 pawn to a queen, but will end up in a losing queen endgame in various versions. So the endgame is probably winning for White, but playing against a prepared "masochist" who also has some two or three hundred Elo points less than you, it makes some sense to deviate earlier.

1 3 . . . �c4 is a rare idea. White is usually advised to play 14 .h4 b5 1 5 .h5 e5 (or 1 5 . . . b4 1 6. til d5 tilxd5 17.�xg7 �xb3 I B .cxb3 til c3t 1 9 .bxc3 bxc3 20.�xc3 �xc3 2 1 .�xc3 Eixc3 22. Eic1 with an advantage in the endgame, Ribli - Velimirovic, Pula 1 97 1 ) 16 .�e3 b4 1 7. til e2 �xb3 I B .cxb3 Eic6 1 9 .hxg6 fxg6 20.�g5! ? with better chances, as in Nikitin - Ignatiev, Moscow 1 963.

13 . . . Eic6!? must be compared with 1 3 . . . �xb3 14.axb3 b5. Now: 1 4.g4 b5! ? ( 1 4 . . . �xb3 transposes to 13 . . . �xb3) 1 5 . til d5 (after 1 5 .!xf6 �xf6 1 6. til d5 �xd2 1 7. Eixd2 Black has 17 . . . �h4!, but interesting is 1 5 .g5 !?) 1 5 . . . �xd2 16 . Eixd2 �xd5 1 7.exd5 with a slight advantage.

14.cx:b3 �k6 1 4 . . . b5 1 5 . Eihe l !was alreadyconsidered

above.

15.g4 1 5 .h4 allows 1 5 . . . h5!?

15 . . . e6 After 1 5 . . . EiacB 1 6.h4 h5!? and now

possible are both 1 7.gxh5!? and 17.g5 til eB I B .�xg7 tilxg7 1 9.f4 (Winants - Sosonko, Dutch Cht 1 992) , where White's chances are at least slightly better. 16.h4 b6

Or 1 6 . . . b5 1 7.a3! (It makes sense for White to prevent 17 . . . h4 and keep his knight on c3 .) 1 7 . . . EiacB (not good for Black are both 1 7 . . . h5 I B .!xf6! !xf6 1 9 .tild5 , and 1 7 . . . b4 I B . til a2! , winning a pawn) I B .h5 e5 (Gruenfeld - Ma.Tseitlin, Israel (Ch) 1 990) and here 1 9 .�e3 looks clearly better for White. 17.g5!? bIgS 18.hxg5 �b5 19.ixg7 �xg7 20.f4


Black's previous play is not too bad, but for the reasons described in the note after 1 2 . .ib3, White's chances can be preferred. 20 ••. �ac8

20 . . . b5? gives White promIsmg attacking possibilities such as 2 1 . � d4 t @gB 22J hh5 gxh5 23.g6!? 2I J�h3!?

2 1 . l':idfl , preparing f4-f5, was also possible as after 2 1 . . . tD g3 White has 22.�h2!.

Also interesting is the immediate 2 1 .f5!? exf5 22.�d4t �e5 23.exf5 1Wxd4 24. l':ixd4 with an initiative in the endgame. 21...YlYc5

Here 2 1 . . .b5 deserves attention. Then White could preserve some advantage by 22.a3! ? �b6 (22 . . . b4 23. tD a2!) 23 .b4 with the idea 23 . . . a5 24.bxa5 �xa5 25 .f5. 22.�d3 b5

This leads to a sharp endgame, which is objectively better for White. 23 . l':ixd6 was not a direct threat, but after passive Black moves White could have improved his position by 23.f5 or 23. l':id4. 23.gxd6 �xd6 24.YlYxd6 YlYxd6 25J'hd6

25 ••• tDxf4 25 . . . b4 26. tD e2± .

26.c!ihb5 �c5? The best chance was 26 . . . a5! . Then

27. l':id7 tDh3! 2B . tDd6 l':ifB looks unconvincing for White as his g5 pawn falls. Better is 27. tDd4, and if 27 . . . l':ihB then 2B.a4 with the ideas: 2B . . . tDh3? 29. tDxe6t! fxe6 30. l':id7t @gB 3 1 . l':idBt @g7 32. l':ixhB @xhB 33.b4+-, and 2B . . . l':ih5 29.b4!. 27.tDxa7 gIgS 28.b4

White's b-pawn is too dangerous. 28 ••• �e5

Or 2B . . . l':igl t 29. @c2 g5 30.b5 g4 3 1 .b6 g3 32.b7 g2 33.bB� l':ic1 t 34. @xc1

60 The Dragon

gl 'lWt 35 . Wc2 and White wins. 29.b5 �xe4 30.b6 �d5 31 .b7 �b4 32.�d7 �f6 33.!k7 �d5 34.�c6! �b6 35J�d7 And Black resigned. 0-1 .

Game 14 van der Wiel - Sax Plovdiv 1 983

1 .e4 c5 2.�a d6 3.d4 cxd4 4.�xd4 �f6 5.�c3 g6 6.Ae3 Ag7 7.a �c6 8.�d2 0-0 9.0-0-0

9 • . . �xd4 The immediate 9 . . . .!e6!? has very

rarely been used by grandmasters in recent years. It is playable to some extent, so we should consider it.

Black's first problem is 10 . Wb l when 1 0 . . . lLl xd4 just transposes to 9 . . . lLl xd4, and it is unlikely that his life is easier in lines such as 1 0 .. J:i:c8 l 1 . lLl xe6 (or l 1 .h4!? 'lWa5 1 2 .lLlxe6 fxe6 1 3 . .!c4

Wf7 14 . .!b3 lLl e5 and now possibly 1 5 . 'lWe2! ? lLl h5 1 6J:i:h3 1:'1xc3 17 .�d2) 1 1 . . .fxe6 1 2.!c4 'IW d7 1 3 .!b3 lLl e5 14 . 'IW e2! , Matanovic - Larsen, Portoroz (iz) 1 958 .

1 0 . lLl xe6 fxe6 i s Black's second problem. His e6 pawn covers d5, but appears to be a weakness itself Now 1 1 . .!c4 'lWc8! 12 . .!b3 lLl a5 may be dubious for Black, but it is at least complicated. l 1 .h4 is uncommon and Black can answer with 1 1 . .. lLl e5 ! . l 1 .g4 (which transposes to the 9 .g4 line) weakens f3 and is therefore rather illogical. White's bishop pair is strong, so 1 1 . .!h6!? may not seem logical either. Still, it is not at all easy for Black to develop counterplay:

1 1 . . . .!xh6 12 . 'lWxh6 lLl e5 1 3 . .!b5 1:'1c8? ! fails to 14 .'lWh3! Wf7 1 5 .f4± Akopian - Alterman, USSR 1 986.

Or 1 1 . . . lLl e5 12 . .!xg7 Wxg7 1 3.f4! . After 1 1 . . . 1:'1c8 12 . .!c4 'lWd7 1 3 .�b3

Black maybe can try to improve on Geller's line 1 3 . . . lLl a5? ! 14 . .!xg7 lLlxb3t 1 5 .axb3 Wxg7 1 6.e5 ± , by 13 . . . lLl h5 , which still looks slightly dubious.

Keeping 1 1 . .!h6 in mind, I also recommend I I .g3 ! ? This is the kind of move which is rarely seen in the Rauzer Attack, but is interesting here as White prepares .ih3 . Black's possibilities are:

a) 1 1 . . . 'lWa5 1 2 . .ic4! . b) 1 1 . . . lLl e8!? 1 2.f4 1:'1c8! (if 12 . . . lLl c7

1 3 .h4!) 1 3 . Wb l (Here 1 3 .h4 lLl b4! 14 . .!h3 Wf7 gives Black counterplay,

but the alternative which I like is 1 3 . i.h3! fi'J c7 1 4. fi'J e2 . 1 3 .i.c4!? can also be considered.) 1 3 . . . fi'J a5 ( 1 3 .. "1M! a5 14. fi'J b5 ! ?) 14 .i.h3 l::ixc3 (not 14 . . . '%!Id7? 1 5 .'%!Id5!) 1 5 .i.xe6t c;!;>hB 1 6.bxc3 fi'J c7 17 .�b3 fi'J b5 and Black's piece play gives him some compensation, 011 - Fedorov, St Petersburg 1 996.

c) 1 1 . . . l::icB 1 2. i.h3 @ f7 1 3 . fi'J e2 '%!1c7 14 . c;!;> b 1 with clearly better prospects for White, Stillger - Farago, Budapest 1 996.

d) 1 1 . . .'%!IcB 1 2.i.h3 fi'J e5 1 3 .'%!Ie2! fi'J c4 14 .i.d4 with advantage, Sonnberger - Stanojevic, Goetzis 1 997.

e) 1 1 . . . fi'J d7 1 2. i.h3! ? l::ixf3 13 .�xe6t c;!;>hB 14 . l::ihfU.

f) 1 1 . . .'%!Id7 1 2 . i.h3 fi'J e5 1 3 .'%!Ie2! b5 1 4.f4 fi'J c4 1 5 .e5 fi'J eB 1 6.�d4 with strong pressure, Romero Holmes -Martin Gonzalez, Linares 1 990.

g) 1 1 . . . fi'J e5 (in some ways this is a principled move) 12 . f4! fi'J eg4 1 3 .i.gl ( 1 3 .i.d4 e5 ! and the queen sacrifice 14.fxe5 ?! i.h6 1 5 .'%!Ixh6 fi'J xh6 1 6.exf6 exf6 17 .i.c4t @ g7 is dubious: Black will transfer his knight to e5) 1 3 . . . l::icB (insufficient is 1 3 . . . '%!Ia5 14 .�c4! Grabarczyk - Jedryczka, Plock rpd 1 994 and 1 3 . . . '%!Ic7 14 . '%!Ie2!? a6 1 5 . i.h3 h5 1 6.i.g2! e5? ! 1 7.h3 fi'J h6 I B .fxe5 dxe5 1 9 .94 Pletanek - Jerabek, corr. 1 999) 14 . c;!;> b l ! ? ( 1 4.'%!Ie2 l::ixc3 ! ? 1 5 .bxc3 '%!I a5 with some compensation, Traub -Bakalarz, Germany 1 99B . 14 .�g2 '%!Ia5 ! ? 1 5 . c;!;> b l l::i c4) 1 4 . . . b5 (or 14 . . . '%!Ia5

1 5 .h3 fi'J h6 1 6. fi'J d5!) 1 5 . fi'Jxb5 fi'J xe4 1 6.'%!Ie2 fi'J gf6 17 .i.h3! ? '%!Id7 I B .�d4 ( 1 B . fi'J d4? ! l::ibB!) I B . . . a6 and now the simple 1 9 . i.xf6 with the idea 1 9 . . . �xf6 20. fi'J d4! is very good for White. On the whole, the entire position after 1 0 . fi'J xe6 fxe6 favours White, but the play is rather strategic in these lines. lO.J.xd4 .le6

A sharp but somewhat unreliable system, which is Black's second most popular option after 9 .0-0-0. 1 1 .bl!

The only advisable alternative to this main move is Timman's 1 1 . fi'J d5! ? , and further lines often can merge with ones which arise after 1 1 . c;!;>b l ! '%!Ic7 1 2. fi'Jd5! ? I t i s easy to see a common idea between 1 1 . c;!;>b l and 1 1 . fi'J d5 - both these moves prevent 1 1 . . . '%!Ia5. 1 l ... 'ilYc7!

Being unable to place his queen on a5 at once ( 1 1 . . .'%!Ia5? 1 2 .fi'Jd5 ±) , Black

62 The Dragon

prepares 1 2 . . .:!HcS and then 1 3 . . . Wla5 - even with a loss of tempo. Not an especially impressive concept, but activating the queen is vital for Black. After other moves he scores terribly:

1 1 . . . �hS? ! is a move which is hard to take seriously. 12 .h4 Wla5 1 3 .g4 gfcS 1 4.a3 (planning 1 5 .h5 and 1 6.h6) 14 . . . h5 1 5 .gxh5 lDxh5 1 6.Wlh6t! �gS 17 . .!xg7 lDxg7 I S .h5 followed by 1 9 . 1D d5 with a decisive attack.

1 1 . . . WlbS?! does not help Black to activate his pieces. 1 2.h4 gcS 1 3 .h5 b5 14 .hxg6 hxg6 1 5 .g4 (it is not necessary for White to accept the pawn sacrifice) 1 5 . . . b4 1 6. lD d5 �xd5 1 7.exd5 a5 I S .g5 lD h5 1 9 . .!xg7 �xg7 20 . .!h3 gc4 2 1 . �g4± Golubev - Glienke, Le Touquet 1 995 .

1 1 . . . WI d7? 12 .�b5! WI c7 13 . .!a4 followed by 1 4 . .!b3 and White is better.

1 1 . . . lD d7 12 .�xg7 �xg7 (Konstantinopolsky - Lisitsyn, Leningrad 1 935) can be met by 13 . lD d5!? preventing 1 3 . . . Wla5 .

1 1 . . . gcS is a natural move, but it does not help Black to develop counterplay. 1 2.g4 (or 1 2.h4 h5 13 .'!e2!? , preparing 1 4.g4, Arakhamia - Matveeva, Jakarta 1 993) 12 . . . .!c4 ( 1 2 . . . lD d7 13 .�xg7 �xg7 14. lD d5! ? lD e5 1 5 . .!e2 with better chances, Golubev - Matveeva, Groningen 1 993) 1 3 .h4 .!xfl 14 .ghxfl (Pachman - Golombek, Venice 1950) 14 . . . gc4 1 5 . Wld3 with initiative.

1 1 . . .a6 just wastes time, as 1 2 . . . b5 and

1 3 . . . b4 is not a serious idea - White can play lD d5 even without invitation. e.g. 1 2.h4 ( 12.g4 is of course also possible) 1 2 . . . h5 ( 1 2 . . . b5 1 3 .h5 b4 1 4. lD d5 , White i s better, Geller - Horowitz, USAUSSR (3) 1 954) 1 3 . lD d5 �xd5 1 4.exd5 Bronstein - Denker, USA-USSR (2) 1 954, and White is clearly better: 14 . . . Wld7 (with the idea of 1 5 . . . Wlf5) is refuted by 1 5 .g4! hxg4 1 6.h5 with a crushing attack. We now return to 1 1 . . . Wlc7.

12.h4 This very direct attempt (White plans

to open the h-file quickly) is my main suggestion for White. After other moves Black faces problems as well. One of these is the positional 1 2. lD d5!? �xd5 1 3 .exd5 gacS 14 . gel ! (defending against 1 4 . . . lDxd5!) . This slightly releases White's pressure, but also keeps Black's queen passive. 14 . . . a6 1 5 .g4 e6 looks more or less playable for Black here. 12 .. JUc8


12 . . . h5 cannot stop White's initiative. 1 3.g4 ( 1 3 . .te2 1'!fc8 14.g4 transposes) 1 3 . . . 1'!fc8 (highly unpleasant for Black is 1 3 . . . hxg4?! 1 4.h5! . Then 14 . . . gxh5 1 5 JWg5! 1'!fc8 1 6 . .td3! gave White a dominating position in Blodstein - Serper, USSR 1 982) , and here:

a) 14.gxh5 provokes the sacrifice 14 . . . Wi a5 , but I do not see any real danger for White after 1 5 .hxg6!? Otherwise, 14 . . . lt'lxh5 1 5 . .txg7 Wxg7 1 6.1'!g l ! Wia5 17. 1'!g5 1'!c5 1 8 .Wid4t f6 19 .b4! Wib6 20. lt'la4! ± Mannion - Pen, Hastings 1 994/5 .

b) 14 . .te2!? Wia5 ( 14 . . . hxg4 favours White. 1 5 .h5 gxh5 1 6.fxg4 Wia5 17.a3! , Mochalov - Roizman, Minsk 1 98 1 ) . Now 1 5 .g5 as in Dominguez - Kudrin, Buenos Aires 2003 does not look like a critical test for Black, but 1 5 . Wi g5 !? is interesting: 1 5 . . . b5 1 6. lt'l d5 .txd5 (or 1 6 . . . lt'lxd5 17.exd5 f6 18 .Wixg6 .txd5 19 .b3!?) 17.exd5 . It is not dear how Black can get sufficient counterplay now. Dominguez gives 17 . . . Wia4 1 8 .Wid2 with a slight advantage to White. In my view, 1 8 .c3 b4 1 9.c4!? , and 1 8 .b3!? Wib4 1 9 .9xh5 It'lxh5 20 . .txg7 Wxg7 2 1 . .td3 also look better for White. 13.b5! ti' as

1 3 . . . lt'lxh5? loses by force after 14 . .txg7 Wxg7 1 5 .g4 It'lf6 16.ti'h6t Wg8 17.e5 dxe5 1 8 .g5 It'lh5 19 . .td3+- Evans -Zuckerman, USA Championship (New York) 1 967. 14.bxg6 bxg6

The relatively little studied alternative 14 . . . fxg6 gives Black more prospects to survive White's kingside assault, but at the same time it worsens the pawn structure. 1 5 .a3 (The immediate 1 5 . .txf6 as usual fails to 1 5 .. J�!:Xc3! ' ) . Now the important difference with 14 . . . hxg6 is that there White's 1 6 . .txf6 .txf6 17. lt'l d5 is not a threat because Black has 1 6 . . . Wixd2 17.lt'lxf6t Wg7! 1 8 . 1'!xd2 Wxf6. But here Black cannot leave the h7 pawn unprotected. He is forced to take on f6 with the e-pawn, creating an isolani on d6. There is no sensible way for Black to avoid White's exchange operation, and it is for White to decide whether he wants to torment Black in an endgame, or to develop an initiative in some different way. Now we consider:

a) 1 5 . . . .tf7 16 . .txf6t or 1 6.g4!? , where Black hardly has anything better than 1 6 . . . . 1'!ab8.

b) 15 . . . 1'!c6 gives White such additional

64 The Dragon

possibilities as 1 6. it'l d5! ? and 1 6 . .tb5!? E:xc3?! 1 7.a4! it'lxe4 1 8 .'lWe l !± Ghyssens - Koller, corr. 1 990.

c) By playing 15 . . . .tc4 Black is aspiring to obtain a tenable endgame. 1 6 . .txf6 .txf6 17.�h6 .txc3!= or 17 . .txc4t E:xc4 1 8 . it'l d5 'lWxd2 1 9 .it'lxf6t exf6 20. E:xd2 E:d8, as in Cichy - Bauer, Bundesliga 1983/4 playoff, may not look too convincing, and White can try 1 6. �el ! ? instead, and i f 1 6 . . . .txfl , then 17 .'IW xfl.

d) After 1 5 . . . E:ab8, 1 6.g4!? can be recommended for White if he wishes to play for an attack. 1 6 . . . b5 (Black creates a threat of 16 . . . b4. The passive 1 6 . . . .tf7 17.g5 ! it'lh5 1 8 . .txg7 it'lxg7 1 9.�h2! it'l h5 20. it'ld5 .txd5 2 1 .exd5 E:ffi 22 . .th3 results in a clearly better position for White, Ivanovic - Kudrin, Lone Pine 1 98 1 ) 1 7.�g5! (probably White also can allow . . . b4 in the lines 1 7.g5 ;!; b4?! 1 8 . it'l b5! and 17 . .txf6 .txf6 1 8 . it'l d5 ;!; b4 1 9 .axb4!? 'lWa4 20.b5) 17 . . . �c7 (After 17 . . . .tf7 1 8 .it'ld5 the continuation 1 8 . . . 'lWd8 19 .�h4! h6 20.g5 hxg5 2 1 . 'lWxg5 ± gives us an idea why g4 can be more useful here than .td3. Also difficult for Black is 1 8 . . . .txd5 1 9 . .txf6, where he cannot allow the white queen to emerge on d5 with check: 1 9 . . . E:c5 20.exd5 ± Lukin - Cebalo, Biel 2004) 1 8 .e5. Now Black's only chance appears to be 1 8 . . . it'le4!? 1 9 .fxe4 dxe5. After the retreat of the bishop from d4, Black will play 20 . . . b4 with an attack. It is a big question whether it gives him

sufficient compensation, but White also has 20 . .td3!? exd4 (20 . . . b4 2 1 . it'l d5!) 2 1 . it'ld5 .txd5 22.�xd5t e6 23.�xe6t 'lWf7 24.�xf7t 'it>xf7 25. E:xh7 with an extra pawn and winning chances, Sebag - Pogonina, Elista 2004 . 15.a3

Black's main idea was not 1 5 . . . E:xc3, which is not killing unless White plays 1 5 . .txf6?, but rather 1 5 . . . b5 ! . 15 ... gab8

Importantly, the immediate 1 5 . . . .tc4 allows White to prepare a brutal, straightforward attack along the h-file by 1 6. E:h3!. Then:

a) 1 6 . . . b5? ! 17 . .txc4 E:xc4 ( 17 . . . bxc4 1 8 . E:dh1 E:ab8 occurred in Short - Velimirovic, Banja Luka 1 985 . Short played 1 9 . 'it>a1 ? ! , but here, in contrast to variations which arise in the line 1 5 . . . E:ab8 1 6 . .td3 .tc4, White has time for 1 9 .'it>c l ! E:b7 20.g4! E:cb8 2 1 . �h2, winning by direct attack - Velimirovic)


1 8J'!dh l ! e5 ( 1 8 . . . b4? 1 9.Axf6 exf6 20. lt'ld5 . 1 8 . . J'!ac8 or 1 8 . . J'!b8 is answered by 19 .Axf6! and 20. i!h7.) 1 9.Ae3 i!xc3 ( 1 9 . . . b4 20.Ah6! bxc3 2 1 . Axg7 cxd2 22.Axf6 and White mates) 20.�xc3 �xc3 2 1 .bxc3 d5 22.Ag5! with a technically winning endgame, van der Wiel - van de Mortel, Wijk aan Zee 1 996.

b) 16 .. . Axfl 17 .i!xf1 i!c4 1 8 . i!fh l i!ac8 (Without . . . b5 this position offers Black more hope. Now he plans 1 9 . . . e5 . ) 1 9 .Axf6!? Axf6 ( 19 . . . exf6?! 20. i!h7!, and following 20 . . . fS funny enough is 2 1 .exfS i!xc3?! 22. i!h8t! ! , but 2 1 . i!xg7t! is also good. After 20 . . . �g5 or 20 . . . g5 White plays 2 1 .�xd6 i!xc3 22.�h2.) 20. i!h7 �e5 (20 . . . �b6? 2 1 . �h6 Wfd4 22.f4! Ag7 23. Wfh2! fS 24.e5! and Black resigned in the e-mail game Schoen -Thorn Leeson, 200 1 in view of 24 . . . dxe5 25 . i!xg7t, etc.) 2 1 .f4 Wfd4 (2 1 . . . Wfa5 ? 22.g4! @f8 23.e5! dxe5 24.g5 seems to be completely crushing.) 22.Wfxd4 i!xd4 23. lt'ld5 i!xd5! 24.exd5 i!c5. Black gets a pawn for the exchange, and some chances for survival. 16.1d3!

van der Wiel's important invention. 16.g4 is less dangerous for Black. 16 ... b5

After this programmed move Black faces problems. The waiting 16 . . . a6 can be met best by 17. i!h4! b5 18 .�g5 and here 1 8 . . . i!c5 19 .ixc5 dxc5 seems to be totally incorrect: 20. �xc5 It'ld7 2 1 . �b4 �c7

(In the case of 17. i!h2?! the white rook would be hanging now.) 22. lt'ld5 ixd5 23.exd5+- Bley - Jackwertch, corr. 1 998.

16 . . . ic4 is often considered to be Black's most realistic chance for acceptable play. The choice between White's main answers is difficult, so we will consider the most important ones:

a) 17.Ae3!? is the move that I analysed in New In Chess Yearbook 11 ( 1 989) . As of now, Black is alive here:

al ) 17 . . . b5? 1 8 .Ah6 Ah8 19 .Af8! is losing for Black: 19 . . . lt'lh5 20. i!xh5 Axc3 (20 . . . gxh5 2 1 .Axe7+- Allemann -Loetscher, Switzerland 2003) 2 1 .�h6!+was pointed out by Oltho£

a2) After 17 . . . lt'le8? 18 .ixc4 i!xc4 19 .1t'ld5 �b5 20.b3! e6, White obtains a big advantage by 2 1 .Ah6! exd5 22.Axg7 @xg7 23.exd5! .

a3) An important line is 17 . . . lt'ld7!? 18 .Axc4 i!xc4 ( 1 8 .. . Axc3? 1 9.Ad4!!) 1 9 .1t'ld5 �xd2 ( 19 ... �b5? 20.b3+- Black

66 The Dragon

loses material.) 20J%xd2 �e8 (20 . . . e6? 2 1 .b3!) 2 1 .!xa7 b6 (This was my main suggestion for Black in 1 989.) 22. �d3 �a4 (22 . . . �c6 23. �hd1 �f8?! 24. �b3 �a8 25 . lt'lxb6! �xa7 26. lt'lxd7t �xd7 27. �b8t wins for White. 23 . . . !f8 gives more chances but is still insufficient.) 23.!xb6 e6 24.!c7 exd5 25. �xd5 It'le5 26.!xd6 It'lc4. Despite White's four pawns for a piece his advantage is not easy to prove (Cordovil - Lecroq, 14th corr. Wch 1 994-2000) .

a4) 1 7 . . . !xd3! ? 1 8 .cxd3 (less ambitious, but interesting is 1 8 .\1�'xd3! ? b5 1 9 . 1t'l a2 Mousessian - Burne, e-mail 2002) 1 8 . . . b5 i s an interesting line. I t may look dubious, even disastrous for Black, but he needs just one move ( . . . b4) to develop serious play, and a forced win for White on the kingside has not been found. 19 .!h6 !h8 20. lt'ld5 (Or: 20.!g5 !g7 2 1 .d4 b4 22. lt'l a2 Wfa4 with counterplay, Ward. Then 23. lt'lxb4 a5 24.!xf6 axb4! 25 .!xg7 bxa3! results in an inevitable draw after, e.g. 26.d5 !? f6! . Also not dangerous for Black is 2 1 .e5 b4! 22.axb4 �xe5 or 20.e5 b4 and 20.!f8 It'lh5! . This was discussed on the ChessPublishing. com forum with an acceptable verdict for Black: 2 1 . !xe7 !xc3 22.bxc3 b4!? , or 2 1 . �xh5 !xc3! 22.bxc3 �xc3 23.Wfxc3 �xc3.) 20 . . . Wfd8 2 1 . !g5 (2 1 . lt'l xf6t !xf6 22.!g5 can hardly give White a significant advantage.) 2 1 . . .lt'lxd5, and now 22. �xh8t �xh8 23. �h1 t �g7 24.exd5 transposes to the game van der

Wiel - Zult, Leeuwarden 200 1 , where Black could save himself by 24 . . . Wf c7!= , with the idea 25.!xe7?! �h8.

b) 17. �h2 ( l7. �h4 or 1 7. �h3 lead to similar play - in the latter case Black obtains the additional resource . . . !e6 at some point.) 1 7 . . . b5 1 8 . �dh1 ( 1 8 .�g5 e5!) 1 8 . . . e5! ( 1 8 . . . b4?! 1 9 .!xc4 and now 19 . . . �xc4? loses to 20.!xf6! ixf6 2 1 . lt'ld5 !b2 22. �h8t, and 1 9 . . . bxc3 20.!xc3 Wfxa3 2 1 . !b3 �xb3 22.cxb3 Wfxb3 23.g4 favours White, Perez -Cabrera, Villa Clara 1 998.) 1 9.!e3 b4 20. lt'ld1 (unimpressive but 20. lt'l a2 !xa2t 2 1 . �xa2 allows 2 1 . . . �c3!) 20 .. . !xd3 2 1 .cxd3 with unclear play after 2 1 . . . � a4! ? De la Riva - Molander, Andorra 200 1 , or 2 1 . . . �b5 22.axb4 E:xb4.

c) 17 .!xc4 �xc4 1 8 .�c1 ! ? (instead, 1 8 .!xf6 !xf6 1 9 .1t'ld5 �xd2 20. �xd2 !;!; gives White slight pressure in the ending, Marjanovic - Messing, Bela Crkva 1 984) 1 8 . . . e6 ( 1 8 . . . �bc8? 1 9 . 1t'l d5 ! �xc2 20. lt'lxe7t �f8 2 1 .Wff4+-. 18 . . . e5? 19 .!e3 and White dominates.) 1 9 .94 �bc8! (Not 19 . . . b5? ! 20.g5 It'l h5 2 1 .ixg7 and now 2 1 . . . �xg7 22. �xh5! gxh5 23. �xf4! with a decisive attack, Bennedik - Demian, corr. 2002. Nor 2 1 . . .lt'lxg7 22.Wff4 b4 23. Wfxd6! �cc8 24. �d5!? Wfc7 25.Wfxc7 E:xc7 26. �b5 ± - Bennedik.) 20.g5 It'lh5 2 1 .!xg7 �xg7 22. �xd6, and here 22 . . . �xc3! (instead of 22 . . . �e5?! 23.Wfd2± Parligras - Cebalo, Nova Gorica 2004) 23.bxc3 It'l g3,


suggested by Ivan Markovic in Chess Informant, leaves White with only a small advantage. Black's main idea is 24. 8:e1 lLle2! 25 . 8:xe2 Wb5t.

d} 1 7.g4

1 7 . . . b5 (Black gets ready to attack by 1 8 . . . e5 and 1 9 . . . b4) 1 8 .!xc4! is, taking into account the absence of a forced win in other lines, my main recommendation for White. Now 1 8 . . . 8:xc4 19 .!xf6 !xf6 20. lLld5 Wxd2 2 1 . 8:xd2 �g7 22. 8:dh2 8:g8 23. 8:h7t �f8 24. lLlxf6 exf6 leads Black to a difficult endgame. 25 . 8:h8!? was played in A. Ivanov - Vigorito, Las Vegas 1 997.

So, he plays 18 . . . bxc4 and the arising position is quite complex. Black, using his major pieces, is able to create some threats against White's a3 and b2 pawns. However, these threats are parried easily, and White's evident positional pluses on three quarters of the board (the cto h-files) should ensure his advantage.

19 . �al ! ? ( l 9. �a2 is of course also possible: 1 9 . . . 8:a6 20.g5!? 8:a6 2 1 .�c1 lLlh5 22. lLl d5 8:b7 23j,xg7 �xg7 24.�e3 Olthof, or 1 9 . . . 8:b7 20. 8:h3!? 8:cb8 2 1 . 8:b l e6 22.g5 lLlh5 23!xg7 �xg7 24. Wxd6 Mishkovski - Niemand, corr. 1 987. Now 24 . . . �xg5 25 .Wh2.) 19 . . . 8:c6 (after 19 . . . 8:b7, 20.g5 ! ? lLlh5 2 1 . !xg7 �xg7 22.�d4t �g8 23. lLl d5 , De Sousa - Molander, Budapest 1 999, preserves White's initiative, but also possible is 20. 8:h3 8:cb8 2 1 . 8:b l ) . Now 20. lLld5? ! Wxd2 2 1 . lLlxe7t �f8 22. lLlxc6 may appear interesting, but it allows Black to equalise by 22 . . . Wxc2!? 23. lLlxb8 Wb3! . Therefore White should not hurry. After, e.g. 20. 8:h3 8:a6 2 1 .�c1 ! ? it is difficult to propose any sensible idea for Black. 17.�g5!

17 .. :�c7?! This is bad, but in the more than 20

years since the game was played Black

68 The Dragon

has not succeeded in finding a reliable antidote to van der Wiel's idea.

1 7 . . . d5?! can be refuted is three ways. The most direct of them is 1 8 . lihd5 j,xd5 1 9 .exd5 b4 20.j,xg6! fxg6 (20 . . . bxa3 2 1 . �h7!+- Turunen - Pyhala, Espoo 1984) 2 1 .�xg6 �a4 22.j,xf6 exf6 23 . Ei:h7 and wins - Oltho£

1 7 . . . j,c4? 1 8 .j,xc4! �xc4 19 . liJ d5 �d8 20.�h4 is just hopeless for Black. Typically for the lines with the bishop on e6.

17 . . . a6 should be met by 1 8 . �h4!? The same advice can be given regarding another waiting move: 1 7 . . . �b7.

17 .. . �xc3! ? 1 8 .j,xc3 �a4 is an exchange sacrifice that does not solve Black's problems, but can be dangerous in practice. So I would ask the reader to pay special attention here. The following variations are given by Nisipeanu and Stoica in Informant 90: 19 . �h4!? b4 20.j,xb4 (not 20.axb4?! liJh7 2 1 . �xh7 j,xc3 22.bxc3 �a3!=, but possible is 20.e5!? liJ d5 2 1 .j,d2 a5 22.j,xg6! liJ c3t 23.j,xc3 bxc3 24. �dh1 �xb2t 25 .1!?c1 +-) 20 . . . liJ h7 (also insufficient is 20 . . . �xb4 2 1 .axb4 �a2t 22. 1!?c1 ) 2 1 . �xh7! I!?xh7 22.�h4t! (an important check) 22 . . . l!?g8 23. �xe7 j,fS (the line 23 . . . a5 24.�a7! �xb4 25 .�a8t! explains White's 220d move) 24.�g5 j,g7 25.f4! with an obvious advantage.

If Black gives away a rook: 17 . . . �c5 1 8 .j,xc5 b4, White needs to make just one precise move, 1 9.j,e3!+-. Then after

1 9 . . . d5 there are many ways, including 20. liJ a2!? bxa3 2 1 .b3 �a4 22. 1!?c1 which is totally convincing.

1 7 . . . �d8!? was recommended by Chris Ward in his book Winning with the Sicilian Dragon 2. There he considers only 1 8 .j,xa7? as an illustrative line. Perhaps the critical continuation is 1 8 .e5 b4 ( l 8 . . . dxe5?! 1 9.j,xe5+- with the idea 20.j,xg6) , where 1 9 .exf6 exf6 is unclear, but 1 9.axb4 is stronger.

Then: a) 1 9 . . . dxe5 20.j,xe5 �xb4 2 1 . j,a6!

(2 1 .j,xg6?! �xb2t 22. l!?xb2 �b6t 23. 1!?c1 fxg6 gives Black compensation.) 2 l . . . �xb2t (2 l . . .�a5 22.j,xc8 liJ d7 is refuted by 23. �xd7! j,xd7 24.f4! j,xc8 25. �xe7+-. 2 l . . .�b6 22.j,xc8 j,xc8 23. 1!?c1 ! ?± .) 22. 1!?c1 ! (even stronger than 22. l!?xb2 �b6t 23.j,b5) 22 . . . �b6 23.j,xc8 j,xc8 and White should win after 24.�d2! or 24.j,d4!.

b) 19 . . . �xb4 20.exf6 exf6 2 1 . �e3!? and


despite the fact that Black is very active, it is not clear how he can obtain sufficient compensation. 2 1 . . . :1'kb8 (2 1 . . Jhd4 22.�xd4 f5 23.�a4! and 2 1 . . .�a5 22. ltJ e4!) 22. ltJ e2!? �a5 23. mc1

17 . . . ltJh7? 1 8 . E!:xh7 �xd4 1 9 .�h6 b4 20. E!:hl �e5 2 1 .f4 results in a decisive material advantage for White. 18.e5! dxe5 19 • .L:e5

White is objectively winning, but he must find a few more good moves. 19 • • • YlYc5

1 9 . . . �b6 20.�xg6! fxg6 2 1 . �xg6 (White is threatening the killing 22. E!:h7) 2 1 . . . ltJ e8 (2 1 . . .�f7 allows the thematic 22. E!:h8t! mxh8 23.�xf7) 22. E!:d6!! exd6 23. �xe6t mfB 24.�xg7t mxg7 25 .�g4t and the game is over - van der Wiel. 20.f4! :ab7

20 . . . E!:b6 does not change much. 2 1 .�xg6 (or 2 1 . E!:h4 �d7 22. E!:dh l Gutman & Reschke, o r 2 1 .b4 �c6

Bennedik - Kreiling, Steinbach 1 998, when White wins with 22.f5 - Olthof, but not 2 1 . E!:h6? ltJg4!.) 2 1 . . .fxg6 22.�xg6 ltJ e8 (22 . . . �f7 23. E!:h8t!) 23.�h7t mfB 24. E!:h5 and White wins easily, e.g. 24 ... �c4 25.�xg7t ltJxg7 26. E!:g5 ltJf5 27.g4 E!:d6 28. E!:e1 . 21 ..hg6! fxg6

2 1 . . .b4 22. E!:h7!. 22.YlYxg6 �

There was no other defence against 23. E!:h7, but now . . . 23J3h8t! I-O

One of the greatest ever Dragon games - especially from White's point of view!

Game 15 Ehlvest - Marin Calcutta 1 997

l.ltJo c5 2.e4 d6 3.d4 cxd4 4.1L1xd4 lLlf6 5.1L1c3 g6 6.0 oig7 7.J.e3 lLlc6 8.YlYd2 0-0 9.0-0-0 d5!

70 The Dragon

Konstantinovsky's move, which in fact is a pawn sacrifice. Nowadays White accepts the offer rather rarely. His choice is still not narrow, but I will consider here only White's main continuation, which was entered into practice by Suetin in 1 955 . 10.exd5 �xd5

Unsuccessful is 10 . . . tLl b4?! , which can be answered by 1 1 ..!c4 tLl fxd5 12 .tLlde2± . 1 1 .�xc6 bxc6 12 . .id4!

White encourages Black either to exchange the dark squared bishops, or to close the al-h8 diagonal. 12 ... e5

The modest alternatives to this move attracted little attention before the 90s. We will consider them in the next two games. By playing 12 . . . e5 Black preserves the Dragon (or, as some say, "Gufeld's") bishop, and hopes to fight for the initiative. From another point of view 1 2 . . . e5 weakens Black's pawn strucrure even more - the d6 square becomes quite sensitive now. 13.J.c5 .te6

Black supports the d5 knight, which is the key detail in his set-up.

After 13 .. J �b8?! the simplest is 14 . .!c4! with an advantage ( 14 .tLlxd5 cxd5 1 5 . 'iNxd5 'IMlf6!? is more complex) . An even more dubious idea is 1 3 . . . e4?! when White can play 14 . tLlxe4!? ( 14.fxe4?! tLlxc3 1 5 .'IMlxd8 .!h6t 1 6.'iNd2 .!xd2t 17J hd2 tLlxe4 18 .i.xf8 tLlxd2;!;) 14 . . . f5 1 5 .tLlc3± .

1 3 . . . ge8!? i s the only real alternative to the main line. White can fight for an advantage in three ways. The current main lines after both 14 . .!c4!? and 14 . tLlxd5 result in slightly worse endgames for Black. I will consider 14 . tLl e4, which is the most principled.

Now 14 . . . .!e6 (which may be best) just transposes to the 1 3 . . . .!e6 line, and thereafter 1 3 . . . ge8 just loses its point.

We will consider two other moves here:

a) 14 . . . f5? ! 1 5 . tLl d6! .!f8!. For a long time this was considered as equalising. In 1 993 Beliavsky introduced 1 6.i.b5! , which was only the first step in the right direction. After 1 6 . . . i.d7 (forced) , White's strong and mysterious novelty 17 .'!a4! {which was found by Deep Blue's support team according to De Firmian, or found by GM Lembit 011 according to other sources, and was played by "Leon" (Shirov) at Internet


Chess Club) , was finally revealed only in De Firmian-Ernst, Stockholm 2002. Black's position becomes strategically dubious even in the case of the strongest 1 7 . . . e4! (Or: 1 7 . . . 1"!:bB I B .c4! , transposing to Misailovic - Jovicic, Tivat tt 1 995, IB . . . tLl b6 1 9.ixb6! YMxb6 20.ib3! and Black will get no real compensation for either pawn or exchange. 1 7 . . . 'iNh4 was mistakenly suggested in New In Chess B/1 99S as winning for Black. White plays I B .ib3! and now I B . . . ih6 19 .ie3 or I B . . . ixd6 1 9 .ixd6 1"!:adB 20.1"!:he l YMxh2?! 2 1 . ixeS ! 1"!:xeS 22J �hl YMg3 23. 1"!:h3. If 1 7 . . . 1"!:e6, then Rogozenko suggests I B .c4 tLl f6 1 9 . tLl b7!? but the immediate I B . tLl b7 may be stronger: I B . . . YMh4 and now, perhaps, 1 9 .'iNaS ! .) .

After 17 . . . e4, amongst White's various promising options there is I B .fxe4! ? (not too bad for Black is I B .ib3 ixd6 19 .ixdSt!? cxdS 20.YMxdSt ie6 2 1 . YM xd6 YM gS t 22.f4 YM xg2;!; De Firmian -Ernst, Stockholm 2002) I B . . . fxe4 19 .ib3 ( 1 9.c4 transposes to Sutovsky - Alterman, Rishon Le Zion 1 994. It continued 1 9 . . . e3 20.ixe3± and White preserves his extra pawn in all variations, but 1 9 . . . ixd6! would be less clear. e.g. 20.ixd6 e3 2 1 . YMe2 'iNaS! ? with the idea 22.ib3 tLl c3) 1 9 . . . ixd6 20.ixd6 and White's positional advantage seems to be quite certain to me, also in the case of 20 . . . e3 2 1 .YMe2.

b) 14 . . . YM c7! ? is a little studied move which I tried against Alexei Shirov in the

Bundesliga. Black's specific idea now is to ensure an early . . . 1"!:dB. I S . tLl d6 looks unclear when the black pawn is still on fl, while after either I s .ic4 or I S .g4 Black can transpose to acceptable lines by I S . . . ie6. I see two principled options for White, where Black in fact has many problems:

b l ) I S .h4. The most direct. I S . . . 1"!:dB (After I S . . . if5 then 16 .tLld6 1"!:edB 1 7. tLlxfS gxf5 can give Black some chances. So White can try 1 6.hS ! ? IS . . . ie6?! transposes to a bad position from the 13 . . . ie6 line where of course White plays 1 6.hS !± . I S . . . f5? ! 1 6. tLl d6 will be followed by tLlxcB and ic4!. I S . . . h6!? just transposes to the main line after 1 6.g4 ie6) 1 6.hS ( 1 6.ic4!? has the idea 1 6 . . . ie6 17. tLl gS! , so Black can play 1 6 . . . f5 1 7. tLl c3 ie6 I B .hS which transposes to 1 6.hS f5 17. tLl c3!? ) 1 6 . . . f5 ( 1 6 . . . if5!? 1 7.hxg6! fxg6 I B .ic4) and now 17. tLl gS allows 1 7 . . . h6!, which looks OK for Black. White must decide between three other quite promising continuations. Both 1 7. tLl c3! ? e6 1 B .ic4! and 17.hxg6!? fxe4! I B .fxe4! ie6! 1 9 .exdS cxdS 20.ia3!? Shirov - Golubev, Bundesliga 2002 leave Black with certain counterchances. White's choice should probably be 1 7.h6!?, which I totally missed in my New In Chess Yearbook 6S analysis. This was tried by Andrei Volokitin in a blitz game against me. ( 1 7. tLl c3!? ie6 1 B .ic4! is very complex.) 17 . . . ihB ( 17 . . . fxe4!? I B .hxg7 ex£3 gives Black some chances,

72 The Dragon

but basically I do not trust his devastated position. 1 7 . . . ifB?! 1 B . lLl f6t± does not work.) 1 B . lLl g5! if6 1 9.ic4! E:bB. Now Volokitin - Golubev, ACP Internet Blitz 2004 continued 20.ib3?? when I could have played 20 .. Jhb3 and 2 l . . . lLl f4+ .

Instead, White must take care of his wonderful bishops: thanks to them, Black is almost paralysed. One idea is 20.a4!? (Not necessarily the best, but 20. E:he 1 and 20. �b1 both allow the remarkable 20 . . . E:b5! ? 2 1 . ia3 E:a5 22.ih4 E:a4, and 20.g4 fxg4 looks unclear.) 20 . . . E:b7 (20 . . . �b7 2 1 .b3. And after 20 . . . f4 2 1 . �xd5t! ? probably only gives a draw - but 2 1 . lLl e4 is good) 2 1 . E:he l �bB 22.b3 with pressure.

b2) 1 5 .id6! ? is a tricky transpositional attempt to avoid the stuff with . . . E:dB and . . . f5 entirely. Then: 1 5 . . . �dB!? (back!) is principled. (Also possible is 1 5 . . . �b6. Now White's bishop can always return to c5, but of course he plays 1 6.h4! first, and after 1 6 . . . E:dB?! ( l 6 . . . f5? ! 17 .ic5! and 1 B . lLl d6. 16 . . . lLle3? ! 1 7.ic5 lLlxf1 1 B . E:hxfl ! ? with advantage) , 17 .ie7! (the main idea behind 1 5 .id6) 17 . . . E:d7 1 B .ic5! destroys Black's scheme. So 1 6 . . . h6 is preferable, hoping to transfer to the 1 3 . . . ie6 main line ( l7.g4 ie6, etc.» 1 6.ia3! ? (claiming that the bishop is placed better on a3 rather than c5) 1 6 . . . �b6!? 1 7.h4 ( l7.ic5 �c7 returns us to the beginning) 17 . . . E:bB 1 B.h5 ifB 1 9 .hxg6. In the notes for New In Chess 1 9 . . . fxg6 undeservedly escaped

my attention, but 20.ic4 ixa3 2 1 . ib3! gives White interesting compensation. Instead, 1 6.ib5!? cxb5 1 7. �xd5 if5! 1 B .�xb5 ixe4 1 9 .fxe4 �g5t 20. �b1 �xg2, as in the computer game Arena -Hagrid, 2002, looks playable for Black. (As I recently found on the web site perso. wanadoo.frllefouduroi, there were two entire thematic computer tournaments on 14 . . . �c7.) It is time now to return to 1 3 . . . ie6.

14 .ixfB? �xfB is premature because of the threat 1 5 . . . ih6! (which also works in case of 1 5 . lLl e4? ih6! 16 .lLl g5 f6) and Black has fantastic counterplay. 14.�e4!

The complex posltlon which arises after 1 4. lLl e4 has been explored quite intensively (636 games in Dorian Rogozenko's recent Dragon CD database certainly means something) . In short, White's basic plan involves ic4 and the advance of his kingside pawns. Fortunately


for the first player Black's most aggressive ideas, linked with . . . f5, do not seem to work well against White's strongest moves: White's only potentially vulnerable point is b2, which can be covered easily even if Black manages to attack it. It is not clear how Black can gain sufficient counterplay to fully compensate for his long-term problems with the pawn structure. 14 . . J�e8

Among the alternatives, only one ( 14 .. J:ibS) has been tried by Black often. Here are all of them:

Black of course would have preferred to put his rook on dS instead of eS, but after the preparatory 1 4 . . 'vY c7 White takes the exchange: 1 5 .!xf8! �xfS ( 1 5 . . . !xfS 1 6.!c4 �bS transposes to 1 4 .. Ji:bS 1 5 .!c4! �c7) 1 6.!c4 with advantage, e.g. 1 6 . . . �bS 1 7.g4 �b6 and now I S .b3!, rather than I S .c3 ttlxc3! with the idea 1 9 .ixe6? ttlxe4 20.!xf7t @hS! 2 1 .fxe4 !h6-+ .

After 14 .. 'vYbS White can play 1 5 .c4!? or 1 5 .!xf8!? (After 1 5 .h4 �dS possible is 16 .!c4 with an already mentioned idea 1 6 . . . 1lYc7 17. ttl g5! , while 1 6.h5 f5 17. ttl g5 e4 looks unclear to me.) 1 5 . . . �xf8 16 .�a5! (the optimal square for the queen) 1 6 . . . �bS (or: 1 6 . . . ttl e3 1 7. �el ! , 1 6 . . . !h6t 17. @b l !e3 I S .!c4±) 17.!c4 �e7! ( 17 . . . f5?! I S . ttl c5 or I S . ttl g5 e4 1 9 .ttlxe6! ? ixb2t 20. @d2 �h6t 2 1 .f4+-) A. Ivanov - Ashley, Philadelphia 1 997. After the 'normal' moves such as I S .h4 Black hardly has full compensation.

Another rare idea is 14 . . . a5, when White has several promising options, the simplest of them being 1 5 .!c4.

Yurtaev's dangerous 14 . . . �bS!? almost certainly should be answered by 1 5 .!c4! ( 1 5 .c4 is another serious move, but the sacrificial I 5 . . . �c7! has not been refuted so far). After 1 5 .!c4! Black has tried various options, but none of them have proved to be satisfactory:

a) 1 5 . . . �eS 1 6.h4 ( 1 6.!xa7!?) 1 6 . . . a5 ( 1 6 . . . f5 is, quite typically, answered not by 1 7. ttl d6? !fS!, but by 1 7. ttl g5 ! e4 I S .c3 which seems to favour White. After 1 7 . . . !h6, .. .f6 is no longer possible and the pin does not trouble White much. After 1 6 . . . h6 1 7.g4! f5? ! I S .gxf5 gxf5, 1 9 . ttl d6 i s already possible, but stronger is 1 9 . �hgl fxe4 20.1lYxh6 �b7 2 1 . �g6! and now 2 1 . . .if5 22. �xg7t �xg7 23.1lYxc6± Zapolskis - Hosruashvili, Leningrad B 1 9S9. Instead, 2 1 . . .�f6?!

74 The Dragon

22Jhf6 i.x:h6t 23. �x:h6 exf3 24. �xe6! �xe6 25 . �xd5! cxd5 26.i.xd5 +- works nicely for White) .

After 1 6 . . . a5 , 1 7.h5!? f5 (analysed by Veselovsky) is rarher unclear. Possible is 17 .i.b3!? (Popovic - Sax, Subotica IZ 1 987) , when 17 . . . f5 is answered by 1 8 . tLl g5 . Also, borh 17.g4 and 17.a4 100k good when rhe almost forced 1 7 . . . h6 can hardly justify Black's previous moves.

b) 1 5 . . . �c7 must be answered by 1 6.i.xf8! i.xf8

17 . 'it> b 1 (Even more principled is 17 .i.xd5! ? cxd5 1 8 .tLlf6t 'it>h8 1 9 .tLlxd5 and White should be better in further complications.) 17 . . . �b6 ( 17 . . . a5!? 1 8 .i.b3 �b7 is a better chance for Black, but still hardly sufficient after 1 9 .c4!?) 1 8 .i.b3 a5 1 9 .c4! tLl e3 20.c5! i.xc5 2 1 . tLlxc5 (2 1 .i.xe6! ? i.d4! 22.i.b3!) 2 1 . . .�xc5 22. �c1 i.f5t 23. 'it>al �d4 24.�xd4! (less clear is 24.�xa5 tLlc2t) 24 .. . exd4 25 .g4 (25 Jhc6 �xb3!)

25 . . . i.d3 26. �xc6 i.e2 27. �hc 1 !± i.xf3 28. �c8t �xc8 29. �xc8t 'it>g7 30. �c7 i.xg4 3 1 . �xf7t 'it>h6 32. �f4! (wirh rhe idea 32 . . . d3 33. �e4!) and White was winning in the brilliantly played game Popovic - Smirin, Moscow 1 989.

c) After 15 . . . a5 ! ? White has usually opted for 1 6.i.b3!? ( 1 6.a4 �c7!? 1 7.i.xf8 �xf8, possibly offers slightly more hope for Black rhan rhe similar line with 1 5 . . . �c7) 1 6 . . . �e8 17.h4, transposing to the Popovic - Sax game.

d) After 1 5 . . . 'it> h8 1 6.h4 (White can also consider taking on f8 or a7, or try 1 6. tLl g5! ? because 1 6 . . . i.h6 now fails to 17 . tLlxf7t) , Black's idea is 1 6 . . . f5 17. tLl g5 i.g8. Still, 1 8 .h5 e4 ( 1 8 . . . �f6? 19 . tLlx:h7 'it>x:h7 20.hxg6t 'it>xg6 2 1 . �h6t 'it>f7 22. �h5! occurred in Sigurjonsson - Kudrin, Gausdal 1 983) 1 9 .i.d4 e3 20.i.xg7t 'it>xg7 2 1 . �d4t �f6 22.h6t 'it>h8 and now 23.i.xd5 i.xd5 24.f4! looks very unpleasant for Black: 24 . . . �xd4 25 . �xd4 �b7 26. �el �e8 27. �e2 �be7 28. 'it>dl 'it>g8 29.b3 'it>f8 30. �d3 i.e4 3 1 . �c3 'it>g8 32. 'it>el �d5 33. �d3 'it>f8 34.c4 i.g8 35 . �c2 c5 36. �d6 �c8 37. 'it>e2 'it>e8 38 .�c3+- and White finally collected rhe e3 pawn in Sammalvuo - Rajlich, Paks 200 1 .

e) 1 5 . . . h6 hardly makes much sense: 1 6.g4 ( 1 6.i.xf8!? , 1 6.i.xa7!?) 1 6 . . . �e8 ( 1 6 . . . f5 17.gxf5 wirh rhe idea 17 . . . gxf5 18 . �hgl 'it>h8 19 . �g6!) 17 .h4 transposes to 1 5 . . . �e8 16 .h4 h6 17.g4.

f) 15 . . . �c8? ! is awkward: 1 6.i.a3!?


( I 6.i.xfB!? '!WxfB 17 .'!Wa5 was the original move order in A. Ivanov - Ashley game - see 14 . . . '!WbB 1 6.h4 E:dB 1 7.h5 lD f4 I B.'!WxdBt '!WxdB 1 9. E:xdBt E:xdB 20.i.xe6 lDxe6 2 1 .hxg6 fxg6!;!; . Also possible is 1 6. '!Wa5! ?) 1 6 . . . E:dB 17 . '!Wa5 with pressure for White, Atlas - Yuneev, Podolsk 1 990.

g) 1 5 . . . f5!? is a desperate attempt to get practical chances. Now 1 6. lD g5 ( I 6.i.xfB!? is another option) 1 6 . . . e4 (after 1 6 . . . i.h6 17 .h4 E:f7 the strongest is probably I B . E:he l ! ? ± - Tiviakov) 17 .�b3 '!Wf6 (if 17 . . . e3?! the simplest is I B.'!Wxe3!) I B .c3 (but not I B .i.d4? e3! 1 9. '!Wxe3 lDxe3 20 . .txf6 E:xf6! 2 1 . i.xe6t E:xe6 22. lDxe6 i.xb2t= Perunovic - Dimitrijevic, Serbia (ch) 2002) I B .. J'!:xb3 (the move order I B . . . e3 1 9 .i.xe3 E:xb3 allows 20.i.d4!) 1 9 .axb3 e3 20.�xe3 lDxe3 2 1 . '!Wxe3 i.xb3 22. E:d7 is a position which I considered in the 199 1 New In Chess article. Then, and now, I cannot see real compensation for Black. 22 . . . c5 , threatening 23 . . . '!Wa6, is answered by 23. E:xa7! .

After the immediate 1 4 . . . f5? ! White has usually played 1 5 . lD g5 i.h6 1 6.h4, which is OK, but 1 5 . i.xfB followed by 1 6. lD g5 is even stronger.

We now return to 1 4 . . . E:eB. 15.h4!

Both 1 5 .c4 and 1 5 . .tc4 allow Black to regroup with 1 5 . . . '!Wc7!. The main aim of 1 5 .h4! is to prevent 1 5 . . . '!Wc7?!, which can be answered with 1 6.h5! E:adB

1 7.hxg6 hxg6 I B .g3 ± Nikulishin - Haba, Voronezh 1 9B 1 .

15 ... h6 The most popular move which can

again be considered as a concession. Black agrees to forget about . . . f5. The alternatives are:

1 5 . . . lD f6?! , unfortunately for Black, simply loses material after 1 6. '!WxdB E:exdB 1 7. E:xdBt E:xdB I B . .te7 lDxe4 1 9.i.xdB i.h6t 20.i.g5 !± .

15 . . . f5! ? 1 6 .lDg5! i.h6 ( I6 . . . '!Wf6 17 .lDxe6 '!Wxe6 I B . .tc4!) is an interesting line. Black is quite active, but his position is too weakened by the pawn advances. 1 7.i.c4 ( I7.c4 is safe and ensures transition to a rather pleasant endgame) 17 . . . '!Wf6 (If 17 . . . lD c7 then a nice idea is l B . '!W d7! ? Baier - Nicholls, corr. 2000. Now could come IB . . . '!Wxd7 19 . E:xd7 E:acB 20. E:xh7! i.xg5t 2 1 .hxg5 i.xc4 22.i.e7! and White is better.) l B . �bl ! E:adB (Sulypa-Golubev, Donetsk (zt)

76 The Dragon

1 998) and here I suggest that White takes on a7: 1 9.i.xa7!? which can be assessed as ± , although untested.

l S .. JJ:b8?! can be answered by 1 6.g4! (or 1 6.i.c4! transposing to 14 .. JJ:b8 l S .i.c4 �e8 1 6.h4) 16 . . . f5 (a logical, but insufficient continuation) 17.gxf5 gxf5 1 8 . 'tJ d6! (also good is 1 8 . 'tJ gS e4 19 .i.d4 e3 20.�g2! 'tJ f4 2 1 . 'tJxe6!, initially suggested by R.Schwarz) 1 8 . . . �f8 ( 1 8 . . . i.f8 is refuted by 1 9.c4! i.xd6 20. �gl t �h8! 2 1 . �h6 �b7 22 . .ixd6 �xd6 23. �xdS! in accordance with Muchnik's analysis) 1 9 .'tJc4! with strong pressure, Timman - Miles, Bad Lauterberg 1 977.

l S . . . aS 16 .ic4 ( 1 6.hS 'tJb4!? with the idea 17.a3? ih6! is given by Tiviakov. The possible improvement for White is instead 1 6.g4!? with the idea 1 6 . . . 'tJ f4 17 .�e1 .idS 1 8 .hS) 1 6 . . . hS ! ? (Ljubojevic's move. If 1 6 . . . a4? ! , White has 17.hS! f5 1 8 . 'tJ d6 �e7 1 9.hxg6 hxg6 20.�gS! , but 1 6 . . . h6 1 7.g4 a4!? makes some sense) . As Black can aspire to reach the position after 1 6 . . . hS in two possible ways (also via lS . . . hS), it deserves double attention: 17.a4 ( 17. g4!? , as in A. Sokolov - Ljubojevic, Bdfon 1 988, is not so clear after 17 . . . 'tJ f4!) 17 . . . �c7 ( 17 . . . 'tJ f4?! 1 8 .�xd8! �axd8 1 9 . �xd8 �xd8 20.ixe6 'tJxe6 2 1 . ib6! with a serious advantage in the endgame, Kudrin - Golubev, Moscow 1 995) 1 8 .g4!? 'tJ f4 1 9 .ixe6 �xe6 and now maybe 20.�d7!? �xd7 2 1 . �xd7 hxg4 22.fxg4 fS 23.gxf5 gxf5. Black's connected

pawns may become dangerous, but after 24. 'tJg3!? �f6 2S .ie3 I would prefer White.

l S . . . hS is "my edition" of Ljubojevic's idea. Now 16 .i.c4 'tJ f4 17.�xd8 �exd8 1 8 .ixe6 'tJxe6 (Howell - Golubev, Bid open 1 993) 19 .ie7! �xd1 t 20. �xd1 'tJ f4 2 1 . 'tJ f6t .ixf6 22ixf6 'tJxg2 leads Black to a dubious but quite puzzling endgame. The most direct 1 6.g4!? may well be the best: 1 6 . . . 'tJf4 ( 16 . . . hxg4 17.hS with an attack, Korneev - Susnik, Kranj 2004) and for example 17 .�el ! ? (Kutuwvic -Baric, Pula open 1 998) 17 . . . .idS 1 8 .c4 (another move is 1 8 .gxhS) 1 8 . . . � c7 1 9 .id6 �b6 20.gxhS .ixe4 2 1 .fxe4 'tJxhS 22.cS �b7 23.ie2 with the idea 23 . . . 'tJ f4 24.hS does not look nice for Black. So 16 .g4 is interesting after both l S . . . aS and l S . . . hS .

l S . . . 'tJ f4!? is Sergei Tiviakov's move, which was topical in the 90s. White is certainly slightly better after 1 6.g3, but I have decided to recommend 1 6.�e1 ! ? , which is more fun - and possibly also the strongest. 1 6 . . . idS ( 1 6 . . . �c7?! 1 7.hS ! ± is given by Rogozenko, whose main explanatory line goes 17 . . . 'tJxhS 1 8 .g4 'tJf4 19 .�h4 hS 20.gxhS 'tJxhS 2 1 . �gS �ad8 22. �xhS �xd lt 23. �xd1 �d8t 24.id3 gxhS 2S. �xhS �e7 26. �h4 �f8 27. 'tJf6!+-.) Now 1 7.hS! brought some fantastic results for White in recent correspondence games. The main direction seems to be 17 . . . f5 ( 17 . . . 'tJxhS? ! 1 8 .g4!? or 1 8 .c4 ih6t 1 9 .i.e3! .ixe3t

20"Ml xe3 � a5 2 1 .cxd5 cxd5 22. 'IMl d2!) 1 8 .hxg6 hxg6 ( I 8 . . . fxe4 didn't serve Black well in Chopin - Hanen, corr. French Ch 2003 after 1 9 .fxe4 �g5 20.�e3 �xa2 2 1 .gxh7t <j(h8 22.g3 ged8 23. gxd8t gxd8 24.b3 'IMlg6 25.�a5 lLl e6 26.�c4 'IMlxg3 27. gel lLl fB 28. <j(b2 with advantage to White.) 1 9 .1Lld6 .ge6 20.g3 gb8! (20 . . . �xf3 2 1 .gxf4 �xhl 22. lLlxf5! .id5 23 .lLlxg7 <j(xg7 24.c4 and White wins - Rogozenko. Also not good is 20 . . . lLl h5? ! 2 1 . lLl xf5 �g5t 22. lLle3 �xg3 23. lLl xd5 cxd5 24. 'IMle3 lLlf4 25 . .ib5 Haugen - Taylor, corr. 2002) .

In this position Black's pieces are active, which gives him many chances. At the same time, White's king seems to feel safer than Black's, so the prospects of the first player can be preferred.

The main potential problem for Black is the emergence of White's bishop on the a2-g8 diagonal after the probable elimination of the black light squared

bishop. Play can continue 2 1 . lLl e4!? fxe4 (Hardly sufficient is 2 1 . . . lLlh5 and now 22.g40r22.�h3 ge8 23 . .ixf5!-Haugen.) 22.fxe4 ge8! (22 ... lLlh5 23.exd5 e4 24.c3 cxd5 25 .�c4! lLlf6 and now 26.�e3 or 26.g4!?) , and here the prophylactic 23.b3!? can be suggested for White (who has at least five or six other possibilities to consider) . 23 . . . lLl e6 (or 23 . . . 'IMlg5 24.�e3! with the idea 24 . . . �xe4?! 25 . .ic4t �d5 26. gxd5) 24.�e3 lLld4 25.exd5 cxd5 26.�xd4!? exd4 27.�d2. White plans to continue 28.�d3 with a lasting positional advantage.

Finally we can go on with 1 5 . . . h6. 16.g4

16 . . . 'i!Yc7 The alternatives are: 16 . . . gb8?! 17.g5! h5 1 8 . .ic4 and

1 8 . . . 'IMlc7 19 .�d6 wins an exchange for White. After 16 . . . a5 17.g5 h5, then 1 8 .a4 (with a probable transfer to our featured game after 1 8 . . . 'IMlc7 19.!c4 ged8 20. 'IMlf2)

7S The Dragon

is more precise in Z. Almasi's opinion than l s .ic4 which allows l S . . . a4! ?

1 6 . . . f5? ! 17.gxf5 gxf5 l S . ltJ d6 followed by 1 9J:J;hg1 is much better for White. 1 6 . . . ltJ f4!? , planning 17 . . . idS is Black's serious alternative. There are other moves, but most often White answers with 17.'�c3 . After 1 7 . . . idS (worse is 17 . . .'�'c7?! l s .id6! or 17 . . . ltJ dS l S .1.Wa3!? with pressure) , White faces an important choice: l S .gS hS! 1 9 .'�·a3 Vlffc7 20. ltJ f6t ixf6 21 .gxf6 E:adS 22.ie7 E:d7 23. E:h2 E:dxe7!? 24.fxe7 Vlffxe7!, Demetrio - Donnelly, corr. 200 1 was examined in great detail in issue 7S of Internet Magazine Co"espondence Chess News, with the verdict of acceptable for Black. l S .hS!? f5 1 9 .9xf5 gxf5 20. ltJ d6 E:e6! 2 1 . E:g1 Vlfff6 22. Vlff a3 E:dS 23. ltJ b7!? E:d7 24. ltJ aS E:eS 2S .c4 ie6 26. ltJxc6 E:xd1 t 27. �xd1 (van Kempen - Gupta, corr. 1 99 1 ) is too complex and risky to be suggested for White, even if he is better here. So I leave it as it is, and go on with I S . Vlff a3! ? Vlff c7 (Black has nothing better) . Now I propose 1 9.hS!? , when 1 9 . . . gS transposes to Psakhis - Vasiukov, Vilnius (USSR Ch) 1 9S0, which is known to be good for White. So, 1 9 . . . gxhS (A cooperative line is 1 9 . . . ixe4?! 20.fxe4 gS 2 1 . ic4 ifS 22.ixfS E:xf8 23.Vlffd6!± , collecting a pawn. After 1 9 . . . f5 20.gxf5 gxf5 2 1 . ltJ d6 E:fS 22. E:g1 �hS White can try 23. ltJc4!? ) 20.gxhS �hS 2 1 . ltJ d6!? (the alternative is 2 1 . E:g1 ifS!? 22. ltJ f6 ixcS 23.VlffxcS) 2 1 . . . E:edS

22. ltJ f5 if6! 23.ia6!? with somewhat better chances. Note that 23.c4? ! .!e6 24.id6? fails to 24 . . . ixf5! 2S . .!xc7 ltJ e2t! 26.ixe2 .!gst. 17.g5!

Another way to fix the kingside pawns: 17.hS gS, now occurs rarely. l S . .!c4 E:edS 19 . Vlfffl as 20.a4 E:abS 2 1 . E:d2 f5 gave Black reasonable counterplay in Ye Jiangchuan - Zhu Chen, Beij ing 1 997. 17 ... h5 18 . .ic4!

l s .id6 Vlffb6 1 9 . .!cS is the typical way to make a draw against a stronger opponent. 18 .. J3ed8

Or l S . . . E:adS 19 .Vlfffl!, and Black's rook on eS is not so useful, while White attacks the a7 pawn already. 19.Vlfffl!

The critical position for 16 . . . h6 (which, at least statistically, is the main line of the entire 9 .0-0-0 Dragon) . White plans to develop his hI rook, and improve his


position step by step: �a3, with a possible invasion of the knight on c5 , is one typical method. Experience shows that it is extremely difficult for Black to hold the position by passive defence. Instead, his only constructive idea is . . . ttl f4, which can be played at once or in the next few moves (after Eid2 and Eihdl it would be too late) . Then . . . ttl f4 is followed by a strategic struggle, where the availability of squares on the d-file for the opponent's pieces is especially important. Black has more weaknesses than White, which ensures an edge for the first player. 19 ... a5

Provoking a4, which will give Black some attacking chances if White later plays carelessly. At the same time . . . a5 allows White to fix this pawn on a dark square, weakens the b6 square and restricts the possibilities of the black queen. The alternatives are:

19 . . . Eid7 20. Eid2 Eiad8 2 1 . Eihdl and there is no obvious continuation of Black's plan (Popovic - Georgiev, Vrsac 1987) . If 2 1 . . .ttlf4 (2 1 . . .a5 22.a4 ttlf4? 23.�b6) , then 22.�xe6!± .

1 9 . . .'i;Yb7 20. Eihe1 ! ? ttl f4 (20 . . . li>h7?! 21 .b3 Wlc7 22.a4 �f5 23.�a3 with pressure, Lupulescu - Golubev, Bucharest 2003. White's main ideas are �b2 and ttlc5, and 23 . . . ttlb6 24.�d6 Wlc8 is answered not by 25.�xf7?! Wlb7, but by 25.Wlc5!) 2 1 .�xe6 ttlxe6 22.�d6 Eid7 (22 . . . ttld4? 23.f4!) 23. Eid2!? Eiad8 24.Eiedl ;!; is akin to the 19 . . . ttlf4 line.

19 . . . ttl f4!? 20.�xe6 ttlxe6 is important.

Not so convmcmg now is 2 1 . �d6 Wlb6!, aiming for 22.c3 Eid7 23. Eid2 Eiad8 24. Eihdl Wlxf2 25. Eixf2 f5! 26.gxf6 �ffi! 27. Eifd2 �h6 Gyimesi - Schutt, corr. 1 997. I vote for 2 1 . Eixd8t!? Eixd8 22. Eidl (not 22.�xa7? Eia8 with the idea 23.�b6 Wlb7! 24.a3 Eib8-+) 22 . . . Eixdl t (22 . . . Eid5?! 23.�xa7) 23. li>xdl and it is not easy for Black to solve his problems. For example, 23 . . . Wla5 24.a3 ttlxc5 25.Wlxc5 Wlxc5 26. ttlxc5 f6 27. ttl e6! fxg5 28.hxg5 h4 29. li>e2 e4 30.fxe4 �xb2 3 1 .a4± Korneev - Komljenovic, Alcobendas 1 994. 20.a4 'l'b7 2Ulhel!?

Apart from this move of Ehlvest, reasonable also is 2 1 .b3!?, preparing 2 1 . . .ttlf4 22.�xe6 ttlxe6 23. Eixd8t Eixd8 24.�b6! as in Z.Almasi - Watson, Bundesliga 1 994/5.

The continuation 2 1 . Eid2 ttl f4! 22.�xe6 ttlxe6 23.�d6 (23 .Eixd8t Eixd8 24.�b6?? Eid8) 23 . . . ttl d4 24. Eihdl Eixd6!

so The Dragon

25 .liJxd6 'IMlb4 26. liJe4 �bS 27.c3 liJb3t 2S . rJic2 'IMlxa4 29. �dSt �xdS 30. �xdSt rJih7 3 1 . rJibl 'IMlal t with a draw in Palac Kolev, Skopje 2002 illustrates what should be avoided. 21 .. J�ab8

After the immediate 2 1 . . .liJf4 22.!xe6 liJxe6 White has 23.!d6!? �d7 (23 . . . liJd4 24.f4! or 23 . . . c5 24.!xc5!) 24.'IMlg3! . 22.b3

22.!b3!? (Ehlvest) is the typical alternative for White. 22 • • • �f4

Two possible waiting moves are 22 . . . rJihS (Marin) and 22 . . . rJih7. In either case quite a logical continuation seems to be 23. �d2 (22.!a3 is also an option) 23 . . . liJ f4 24.!xe6 liJxe6 25.!d6!? Now either version of Black's exchange sacrifice (25 . . . �xd6 or 25 . . . liJd4) would hardly work well for him, while after 25 . . . �bcS White will at least have 26. liJ c5! ? with a positional advantage. 23.ixe6 �xe6

24.�f6t!? Such an exchange of the knight for

the bishop is always a major decision for White. Even if he wins the e5 pawn (which is usually the aim of liJf6t) , Black can sometimes organize counterplay, using his queen and knight duo. But, importantly, here the black queen cannot be activated easily.

Also possible is 24.!d6! ? 24 ... !xf6 25.gxf6 gd5

25 . . . �xd lt 26J !xdl :BeS 27.'IMle3!? and White is better. 26 . .id6 E:d8?!

Both players mentioned 26 . . . �cS!? 27.!xe5 c5 in their annotations. I believe Black cannot get full compensation after 2S. �xd5 'Mfxd5 29.c4 'IMlb7 30.'Mfe3. 27.ixe5!

Better than 27. �xd5 cxd5 2S.!xe5 d4! . 27 ... 'ilYb4 28.E:xd5 cxd5 29.J.b2!

29.'Mfd2 'Mfxh4 30.'Mfxa5 d4! ? gives Black counter chances. 29 ... d4

After 29 . . . 'Mff4t 30. 'Mfd2! 'IMlxh4 3 1 . 'Mfxa5 (± Ehlvest) , 3 1 . . .d4? is impossible because of 32. �xe6 fxe6 33.'MfxdSt. Marin decides to advance the d-pawn immediately, but faces another problem: 30.gxe6! Instead 30.'IMld2 cJth7 3 1 . 'IMlxb4 axb4 32. �e5 g5 33.hxg5 rJig6 gives Black counterplay (Marin) . 30 ... fxe6 31 .'ilYg3!

Now Black cannot defend his g6 pawn with the king because of 32. Wi cn, so White wins it, obtaining a technically won position. 31 ... �f8

If 3 1 . . . <ti>h8?! 32. <ti>b1 W!b7 33.W!xg6 W!h7 34. Wi g3! gc8 3S . WigS ! Wixc2t 36. <ti>a2 W!h7 37.,ixd4 followed by W!xaS with an easy win. 32.�bl §b7!

Not 32 . . . d3? 33. W! c7 dxc2t 34. <ti>a2! c1 ctJt 3S. <ti>b1 and Black can resign. 33.§xg6 §f7 34.i.a3t �e8 35.§h6

Also possible was 3S . W! gS (Ehlvest) 3S .. Jl:dS 36. Wif4 <ti>d7 37.,ie7, etc. 35 ... e5

Hardly better is 3S . . . d3 36.cxd3 eS (or 36 . . . gxd3 37.,ie7! gd1 t 38. <ti>b2 ggl and now 39. <ti>c3!? , avoiding the checks) 37. <ti>c2, etc. 36.§g5! gd5 37.§f5! �d8 38.f4 d3?

Black could have continued his unpromlsmg defence by 38 . . . exf4 39.,ie7t <ti>c7 40.W!xf4t, etc. 39.cxd3 gd4?

The final blunder in time trouble. 40.,te7t �c7 41 .§xe5t 1-0

Game 16 Balinov - Velickovic Vienna 1 998

1.e4 c5 2.�a d6 3.d4 cxd4 4.�xd4 �f6 5.�c3 g6 6.J.e3 J.g7 7.a 0-0 8.§d2 �c6 9.0-0-0 d5 10.ad5 �xd5 1 1 .�xc6 bxc6 12.id4 ixd4

This is less ambitious in comparison with 12 . . . eS. Now Black is essentially fighting only for a draw. 13.§xd4 §c7

This move of Hodgson is stronger than 1 3 . . . W!b6 14. ctJ a4!. This practically wins a tempo for White because c3 is not an optimal square for his knight. 14 . . . W! as ( 14 . . . Wixd4 l S . gxd4 is just what White wants) l S .b3! ,if5 ( 1 S . . . gb8 1 6.W!cS! , forcing the exchange of queens was good for White in Suetin - Vasiukov, Voroshilovgrad 1 9S5) 16.WicS (a serious, but much more complicated alternative, is 1 6.g4 with the idea 16 . . . ixc2 17 .gd2) 1 6 . . . WixcS 17. ctJxcS ctJc3 (or 17 . . . ctJ e3 1 8 . ge1 and Black is suffering) 1 8 .ge l ! (not 1 8 . gd4?! gfd8 , and after 19 . 9xd8t gxd8 20.g4 gdS! Black is fine) 18 . . . ctJxa2t 19 . <ti> b2 ctJ b4. Here I had the "pleasure" to meet the novelty 20.,ic4! over the board. (20. gxe7?! gfe8! , Korneev - Summercale, Groningen 1 995 , with the idea of2 1 . gb7 ctJdS 22.,ic4? ic8! and White's rook is trapped.) 20 . . . gfe8 2 1 .g4! Here we play for two results, and a zero for White is not among them: 2 1 . . .ic8 (similarly discouraging is 2 1 . . .ixc2 22. ctJ a6! -van der Wiel. 22 . . . ctJd3t 23. <ti>xc2 ctJxe l t 24. gxe 1 where Black must follow with 24 . . . gac8 and await his fate) 22 . geS!? , etc. van der Wiel - Golubev, Germany 1 999. 14.§c5

I cannot expect that every reader will have time to study two or more complex

82 The Dragon

directions in full detail. So this move, as a result of a hard and uncertain choice, is my main recommendation for White.

14 .i.c4, twice used by Shirov against Fedorov, is no weaker but more complex. Then 14 . . . lLl b6 is Black's most popular move, when White can fight for the initiative in various ways. Shirov played 1 5 .i.e2!? ( l 5 .i.b3 c5!), trying to underline the drawbacks of Black's passive knight retreat. Alternatively, White can allow . . . lLlxc4, because such an exchange cannot be called a strategic achievement for Black. Instead of 14 . . . lLl b6, Black can play actively with 14 . . . e5!? 1 5 .�d2 (after 1 5 .�c5?! lLlxc3, 16.�xc3 is no longer possible) 1 5 . . . i.e6. Here 16. lLl e4 gad8! ? (after 1 6 . . . gab8 17.h4 Black has more problems) 17 . lLlc5! ? looks slightly better for White.

White's reasonable options also include the rare moves 14.h4 and 14.g3! ? 14 . . . tihc3

The principled alternative is 14 . . . �f4t. Kiril Georgiev successfully used this move against Almasi in 1 996, but did not repeat it against Morozevich in 2000. 1 5 . @b1 i.f5. Now White can obtain a relatively small advantage in two ways: 1 6.i.d3 lLlxc3t ( l 6 . . . lLl e3!? 1 7.g3 �xf3 1 8 .i.e2 i.xc2t 1 9 .@c l ! i.xd1 , and now possibly 20. gxd l ! ? �f2 2 1 . gfl �xh2 22.�xe3 with an initiative.) 1 7.�xc3, and 1 6. lLl e2! ? .

He can also take on d5, which involves some risk but seems to be the most principled: 16 .lLlxd5 cxd5 17 .�xd5 (not 17.�xe7? i.xc2t 1 8 . @xc2 �a4t 19 . @d2 gac8! 20.i.d3 �a5t 2 1 . @e2 gfe8-+ Alterman) 17 . . . gab8 and here:

a) 1 8 .�d4? is refuted by 1 8 . . . i.xc2t! (Alterman) 19 . @xc2 gfc8t 20. @b1 gclt .

b ) After 1 8 . gd4?! Black gets good play by 1 8 . . . �c7 or 1 8 . . . �e3 - Alterman.

c) 1 8 .i.c4 �c7! ( l 8 . . . i.e6?! 19 .�d4

Experts vs. the Sicilian S3

�xd4 20. E(xd4 E(b4 2 1 .b3 i.xc4 is refuted by 22.a3 !± Balje - Nagley, corr. 1 999) 1 9 .i.b3 ( 1 9.b3?? i.e6) 1 9 . . . E(bdS!, and here a draw was agreed in Almasi -Georgiev, Cacak 1 996. White could fight for something by 20. �b5 E(xd1 t (20 . . . �xh2 2 1 .g4 i.cs 22. E(de l !?) 2 1 . E(xd1 �xh2 22.g4!? , but it looks rather double-edged.

d) I do not see a convincing way for Black to obtain full compensation after l S .b3!?

e) Worthy of attention is l S .�d2!? For example, IS . . . � e5 (An alternative is lS .. . �a4 1 9.i.d3 E(fdS 20. E(he1 i.xd3 2 1 .cxd3 when it is not so easy for White to convert his extra pawn.) 1 9 .b3 E(fdS ( 1 9 . . . a5 20.g4! i.e6 2 l .f4 followed by 22.f5!.) 20.�xdSt E(xdS 2 1 . E(xdSt c;tg7 and now not 22.i.d3 �a5! but 22. E(d1 ! . The presence of bishops improves Black's chances in such endgames, but White's extra pawn is still an extra pawn.

16.M!? White is aiming to create weaknesses in

Black's kingside. 16 . . JUd8

1 6 . . . i.xa2?! 1 7.b3 a5 l S .h5! favours White. For example l S . . . g5 19 .h6 e5 20 . .tc4 and after 20 . . . a4 2 1 . E(d7! �xd7 22JWxe5 Black is mated on g7. 17.i.d3 c5 18.h5 �f4t 19.�bl �d4 20.bxg6 bxg6

15.�xc3 .le6 \III 21 .sa5! Black's king is potentially vulnerable

and White rightly avoids the exchange of queens. With the same idea, weaker would be 2 1 . � e 1 (Kurnosov - Solovjov, St Petersburg 2004) because of2 1 . . . E(abS 22.b3 c4! . 21 . . J�ab8 Not 2 1 . . .�b4? 22.�c7! . 2 1 . . . E(d7 22. E(de 1 could transpose to the game after 22 . . . E(bS 23.b3. 22.b3 gd7!

It is hard to propose a better move: 22 . . .

84 The Dragon

c4? 23.i.e2 22 . . . E1d6? 23.1Wxa7. In the game Black wishes to exchange queens by 23 . . . �b4. 23.E1del! �d5!

23 . . . �b4 loses a pawn after 24.1Wxb4 cxb4 25 .i.xg6!, as well as 23 . . . c4 24.i.xc4! i.xc4 (24 . . . �xc4? 25.1We5) 25. E1e4. 24.�e4!

24 ... e6 The alternative 24 . . . E1bd8!? 25 .i.xd5

1Wxd5 would hardly resolve all Black's problems: 26. E1dl ! 1Wg5 (or 26 . . . �xdlt 27. E1xdl E1xdl t 28. @b2, winning one of Black's pawns) 27. E1xd7 (rather than 27.�xd8t E1xd8 28. E1xd8t @g7 29.g4 c4!?) 27 . . . E1xd7 28.1Wa6!?, and it seems that after 28 . . . E1d8 (28 . . . e5!?) 29.�xa7 �xg2 30. E1el 1Wxf3 3 1 .1Wxc5 e6 32.a4 White's pawns should be faster. e.g. 32 . . . E1d5 33. 1Wc7 1Wf2 34. E1hl @g7 35 .1Wc3t! 1Wd4 36.1Wxd4t E1xd4 37.a5 f5 38. @b2, etc. 25.ixd5 exd5

Better than 25 . . . �xd5?! 26. E1e4± (or 26. E1dl !?) , or 25 . . . E1xd5 ?! 26.1Wxa7. 26J�Ml!

White removes his rook from one of the open files that he controlled - it is more important to disturb Black's centralised queen! 26. E1h3 would have been premature as 26 . . . 1Wb4! forces a queen exchange. 26 ... �b4

27.VNa6! Black should have been able to defend

the endgame after 27.1Wxb4 cxb4!, but not 27 . . . E1xb4 28.c4 d4 29. E1he l ! . 27 ... VNb6?

Only here does Black go clearly wrong. 27 . . . @g7? loses to 28. E1h7t!! @xh7 29.�f6. After 27 . . . 1Wc3 White wins the pawn by 28.1Wc6 c4 29. E1xd5. The continuation 27 . . . 1Wf4 28. E1h3!? (28 .1Wc6 �d6 29.1Wxd6 E1xd6 30.c4;!;) 28 . . . c4 29. E1dhl 1Wd4 30.1Wc6 E1bd8 allows Black to


maintain the material balance, but having passive rooks and an unsafe king his chances for survival are uncertain. 28.'ilYa4!

A kind of fork: the d7 rook is attacked and White gains time to transfer his queen to the kingside. 28 •.. 'ilYd8

Otherwise 29 .'�rh4 would have been decisive. 29.'ilYf4!

With two threats: 30.11Nh6 and 30. �h2. White's attack seems to be unstoppable. 29 •• J:�b4 30.'ilYe5

Even simpler was 30.11Nh6! �f6 3 1 . �de l !+-. 30 .•• f6 31.'ilYe6t �g7 32.'ilYh3!

Black could continue his suffering for a few more moves, but he blundered his rook by 32 ... 'ilYg8��

and immediately resigned. Still, the game was very well played until Black's understandable mistake on the 27th move. It gives us an example of White's optimal strategy in positions with Black's weakened queenside pawn structure: to open a second front on the kingside, in accordance with the "Principle of two weaknesses" . 1-0

Game 17 Rowson - Mah Birmingham 1 999

l .e4 c5 2.tilO d6 3.d4 cxd4 4.tilxd4 tilf6 5.tilc3 g6 6.J.e3 J.g7 7.0 0-0 8.'ilYd2 tilc6 9.0-0-0 d5 10.exd5 tilxd5 1 l .tilxc6 bxc6 12 . .id4 tilxc3!�

This move is linked with a similar concept to 1 2 . . . ixd4: Black limits his ambitions and goes for exchanges, aiming to defend a slightly worse position. 13.'ilYxc3 .ih6t!

1 3 . . . ixd4?! 14. �xd4 �b6 and now 1 5 .h4!? is the most direct. 1 5 . . . ie6 ( 1 5 . . . h5 16. �e4 ie6 17.g4 id5 18 . �e3 hxg4 19 .h5 with an attack, Brkic - Baric, Bizovac 2003.) 1 6.h5 �fd8 17 .�dh4!? ( 17. �xd8t �xd8 1 8 .hxg6 forcing the weakening 1 8 . . . fxg6 is also interesting.) 17 . . . g5 1 8 .h6 f6 19 . �e4 with a dangerous initiative, Linford - Pym, England 2003. 14..ie3

14 . �b 1 ?? loses the bishop after 14 . . . e5 ! . 14 ... he3t 15.'ilYxe3 'ilYb6!

The best chance.

86 The Dragon

Worse is 1 5 .. .v��' a5 ?! 16 .�c4 with a strategically dominant posltlon: 1 6 . . . �a6 17 .�b3! e6 ( l7 . . . c5 fails to 1 8 J �d5, Alterman & Vaisman) 1 8J �d6 c5 1 9 .'Wel ! ? (I. Almasi's suggestion, which forces an exchange of queens. Also good is 1 9.'We5 :Eiad8 20. :Eihd1 'Wc7, Isupov - Maksimenko, Vladivostok 1 995 . Now 2 1 . :Eixe6! :Eixd1t 22. mxd1 is recommended by Olthof.) 1 9 .. :�xe 1 t 20. :Eixe 1 �b7 2 1 . :Eid7!? and it is hard for Black to save his pawns.

Similarly unadvisable for Black is 1 5 . . . 'Wc7?! 1 6.�c4! and now, for example, 1 6 . . . :Eib8 17.h4!? �b6 1 8 .�c3. Here Black is a tempo down compared to the normal lines. 1 8 . . . �b4 (or 1 8 . . . �a6 19 .�b3!? c5 20.h5 c4 2 1 .hxg6 �xg6 22.�xc4 :Eifc8 23.�d3 with an extra pawn) 19 .�xb4 :Eixb4 20.b3, and in this endgame Black will suffer greatly.

The idea of offering a pawn sacrifice by 1 5 . . . 'Wb6 attracted attention in 1 990,

after Boris Alterman employed it in a game against Sergei Tiviakov. This stem game was followed by active discussion. Later, Kasparov's win with White over Topalov in 1 995 delivered a psychological blow to the supporters of this line. The objective theoretical evaluation of the line is similar to that of 12 . . . �xd4: a slight advantage and almost no risk for the first player. 16.YlYxe7!

Accepting the offer. Black now obtains a slight initiative, which does not fully compensate for the pawn. Less principled is 1 6.'Wc3!? �e6 17.h4 :Eifd8 1 8 .�d3 with a small plus, when 1 8 . . . 'We3t 1 9 .mb 1 c5 20.h5 � d4, which transposes to Balinov - Velickovic, does not look like Black's best choice. 1 6. 'W xb6 axb6 (Tiviakov - Alterman, Sochi 1 990) 17�c4!? b5 1 8�b3 c5 1 9�d5 :Eia7 is acceptable for Black according to Alterman & Vaisman. 16 . . .i.e6!

Experts vs. the Sicilian B7

16 .. J:ibB puts little pressure on White. 17 .'�f6 �e6 (If 17 .. .'�f2? I B .'�d4! forces an exchange of queens, and 17 . . . �f5 is parried by I B .�d3!') I B .'lWc3! l"!fdB 19 . 1"!xdBt l"!xdB 20.�d3 (20.h4!?) 20 ... 'lWf2 2 1 . 'lWd2 'lWd4 22. l"!dl l"!d7 (not 22 . . . �xa2? 23.�e4! andWhite iswinning) 23.a3 . Here the draw was agreed in Brod Bonstingl, Gamlitz 1 99B. Perhaps White was worried that Black would be able to claim compensation after 23 . . . 'lWd5, but the continuation 24.h4 'IW a2 25 . 'IW c3 clearly favours White.

After 1 6 . . . �e6 Black's main threat is 17 . . . 'IW e3t! with a guaranteed draw following I B . l"!d2 ( I B .<j(b l ?? �xa2t) I B . . . 'lWe1t . 17.YlYa3!

The most solid continuation.

17 •• .lUd8 The alternatives include: 17 . . . l"!abB can be answered well by

I B .id3 .

17 . . . a5 I B .�d3 'lWb4 19 .ie4 (Not 19 .'lWxb4 axb4 20.a3 bxa3 2 1 .b3 c5, preparing . . . c4.) 19 . . . 'lWxa3 20.bxa3 l"!a6 (Schneider recommended 20 . . . l"!abB 2 1 . �xc6 �xa2 22.�e4! l"!fcB but after 23. <j(d2! Black has no compensation) 2 1 . l"!d6 l"!cB 22. l"!hd 1 . White is likely to exchange bishops by 23.�d5 on the next move, and Black's life will be hard. GM Tolnai tried to defend the Black side as many as three times, and his two draws with one loss looks like quite a good result.

17 . . . 'lWf2, and after I B .h4!? l"!fdB 1 9.�d3 'lWxg2 20.h5 'lWxf3! Black probably should be able to defend his king in further complications. Instead, I B .'lWa5 , taking control of dB, is a common move for White. After I B . . . l"!abB ( I B . . . �f5!? 19 .'lWd2 'lWc5 20.�d3! (Pupo considers 20.'lWc3 'lWxc3 2 1 .bxc3 ie6 22.c4 t, which does not look like a winning endgame) 20 . . . l"!fdB 2 1 . <j(bl �xd3 22.cxd3 and Black has problems) White as a rule has opted for 1 9.h4. Here it seems that Black has to play 1 9 . . . �f5! . White's advantage does not look impressive here. 20. 'IW d2 'IW c5 and Black's very strong threat now is 2 1 . . . l"!xb2! ' But 19 .h4 is not necessarily best. One alternative is 1 9 .b3!? and after 1 9 . . . 'lWe3t?! 20. <j(b2 followed by 'lWc3, and 19 .. . if5 20.'lWd2 'lWc5 2 1 . �c4 Black does not have full compensation.

17 . . . l"!adB! ? (initially suggested by Kasparov) may be slightly more precise

88 The Dragon

than moving the other rook. 1 8 .�d3 �d5 (According to Tiviakov, insufficient is 1 8 . . . �f2 19 . �d2 �e3. White can continue with 20. �hd1 .) and we come to a very important position.

a) 1 9 .'�c3 �c5 20.�f6 �xa2 looks playable for Black: 2 1 .h4 (or 2 1 . �he1 �d5, planning 22 . . . �a5 , Pelletier -Berndt, Germany Bundesliga 1 999) 2 l . . . �b8!? (after 2 l . . . �a5 White can try 22.b3!?) 22.h5 (22. �he1 ! ? can lead to a total mess, e.g. 22 .. .'�a5 23.h5 �xh5 24.b3 �c5 25 .�a1) and 22 . . . �xh5! equalizes.

b) 1 9 .b3 gfd8 will be considered via the move order 17 . . . �fd8.

c) I suggest 19 . �he1 �a5 20. �c3 �xa2 2 1 .b3 �a5! 22.'it>b2!? (Rogozenko) , with a slight advantage. Note that 2 1 . . .�a5 ?! 22. �xa5 �xa5 and now, according to Rogozenko 23. �b2 gives an unpleasant endgame for Black. 23 . . . �h5 24.h3 a5 and now 25 . �e4 �d8, Kasimdzhanov

- Alterman, Bad Wiessee 1 997 where White has 26. �de 1 ! ? - Rogozenko, or perhaps even better is 25 .�e4!? 18.i.d3

After 1 8 .�a6!? (Kasparov - Topalov, Amsterdam 1 995) Black can try 1 8 . . . �ab8!? 19 . �he1 �d5 20. �xd5 cxd5! or 18 . . . �d5!? 19 . �xd5 cxd5! .

18 . . J:�d5 After 18 . . . c5, 19 .�e2!? transposes

to the Kasparov - Topalov game, which continued 1 9 . . . c4 20.f4! �d4 2 1 . �xd4 �xd4 22.g3± . Instead, 1 9 .h4 c4 20 . .te4 �ac8 2 1 .h5 (Arnold - Bozinovic, Vienna 2003) 2 l . . .f5 ! ? is complex and, most likely, not bad for Black. 19.9he1!?

White proceeds by analogy with the Kasimdzhanov - Alterman game.

Other approaches are: a) 19 .�c3 �c5 20.�f6 �xa2 2 1 . �he1

� a5 ! ? and Black i s probably alive. After 2 1 . . . .td5 ?! the small difference with the

Experts vs. the Sicilian S9

Pelletier - Berndt game allows White to win a crucial tempo: 22. E:e7 E:a5 23.b3 �b4 24. E:de l ! E:f8 25 . E:eS! with advantage, Flores - Ballesteros, corr. 1 997.

b) 1 9 .b3 E:adS 20. E:he1 a5 (maybe Black could try 20 .. !f5!?) 2 1 . �e7 E:Sd7 (or 2 1 . . .a4 22. E:xe6 fxe6 23.�xe6t @g7 24.'�e7t @h6 25. E:el ) 22.�f6 a4 23 . E:e4 axb3 24.axb3 and White's chances are preferable, Furlan - Gomboc, Ljubljana 1 995.

c) Curious is also 19 . E:de l ! ? (as in van den Doel-Zomer, Vlissingen 2002) . 19 •• J�a5 20.%Vc3 l:i:xa2 21 .b3 c5?

A careless move, which will be refuted neatly.

Akin to Kasimdzhanov - Alterman is 2 1 . . .�a5?! 22.%Vxa5 E:xa5 23. @b2 with an initiative in the endgame.

Better is 2 1 . . . E:a5 22. @b2 with a slight advantage for White. If 22 . . . E:bS (threatening to win by 23 . . . E:c5!) , then

23. E:xe6!? (not 23.!xg6? E:c5 24.�d3 E:d5!) 23 . . . fxe6, and here 24.!c4 E:d5 25.!xd5 cxd5 26. E:al ;!; preserves White's pluses, while one may also consider 24.!xg6 E:d5! (24 . . . hxg6?? 25 . E:d7 e5 26.'IM!'d3!) 25 .!d3!? 22.gxe6!!

Here this sacrifice allows White to use the awkward position of the black rook. 22 • • • %Vxe6

After 22 . . . fxe6?! 23.!c4 White completely dominates. 23 • .ic4 %Vc6?

23 . . . %V cS was the only way to avoid the immediate loss: 24. @bl E:a6 25 .!xa6 (25 .E:e l ! ? with the threat 26. E:e7 can be inserted) 25 . . . %Vxa6 26.�xc5 �e2 27. � d5 E:cS 2S. E:d2 with a healthy extra pawn for White.

23 . . . %Vb6? would save a rook, but lose the king: 24.!xf7t! @xf7 25. E:d7t @e6 26.�g7 E:al t 27. @b2. 24 • .id5! %Va6 25 • .ixa8 galt 26.<bd2

Certainly not 26. @b2?? �a3 mate. 26 ••• %Vd6t 27.%Vd3 %Vf4t 28.<1>e2

and Black admitted defeat. 1-0

90 The Sveshnikov

The Sveshnikov

- By Jacob Aagaard

It is almost impossible to describe the huge changes the Sveshnikov has undergone since I wrote a small book on it for Cadogan in the late 90s. A great contribution to this opening has been delivered by players such as Kramnik, Leko, Kasparov and Shirov, as well as lesser known grandmasters, but still experts on the opening in their own right, like Rogozenko (who published a CD for ChessBase with the opening, as well as writing various articles for different magazines) , Yakovich (who wrote a book on the opening for Gambit, which is great despite some flaws) , McShane, and the latest addition, the world's youngest grandmaster, Magnus Carlsen. To give a strong recommendation against an opening that is favoured by such a group

of outstanding grandmasters is by no means easy. Still it is possible to give some useful practical advice on where to look for an advantage and some indication of where the most recent successful assaults on this solid defence have been made.

The main line I have chosen against the Sveshnikov { 1 1 .c3 and 1 2.ex5) is in many ways the most practical, as well as being objectively a strong line, as it does not allow Black to choose between two main lines, as he can against 1 1 ..td3. The main game, Hector - Carlsen, clearly proves that Black needs to find a different way to treat this position, as the very simple harmonious set-up demonstrated by the Swedish grandmaster brought the Norwegian boy wonder real troubles. The solution chosen by Carlsen was a desperate bishop sacrifice, which ultimately brought him the draw, but white's play could be improved.

After 1 .e4 c5 2.tLla tLlc6 3.d4 cxd4 4.tLlxd4 tLlf6 5.tLlc3 e5 6.tLldb5 d6 we have the standard position of the Sveshnikov Sicilian. Here the main choice for a long time has been 7.J.g5 a6 8.tLla3 b5 when White can either choose 9 . lt'l d5 or the sharper move, creating weaknesses on the Black kingside: 9.J.xf6 gx:f6 IO.tLld5 f5

In this position Black could also play 1 0 . . . .tg7 with the idea of It'l c6-e7, but if White answers l 1 .c3 then Black cannot avoid transposition, as his only fully playable move is 1 1 . . . 5 .


Now after 1 1 .c3 J.g7 12.exf5 .ixf5 13.�c2 Black can choose between different ways to play this position. There is 1 3 . . . i.e6 with the idea of a quick lZl c6-e7 to exchange a knight on dS . White will in this case play 14.g3! , a move made main line. The idea is to recapture with the bishop instead of the knight on dS, as the exchange of bishops would favour White; partly because it eliminates the bishop pair, but also because of light squared weaknesses in the Black camp, and because the white knight would do little good on dS.

Usually these days Black does not play this but 13 • • • 0-0 14.�ce3 J.e6. Here 14 . . . i.g6!? is an interesting alternative, popularised by Leko and not so easy to meet. The main point is I S .h4 i.e4! when Black has provoked White into weakening his kingside. However, the main line is still 14 . . . i.e6, when I suggest following continuation: 15.i.d3 e; 16.0-0 ga7 17.a4!

Theory does not consider this troublesome for Black - yet. However, once the gravity of the Hector game becomes apparent, Black players will most likely look elsewhere for satisfaction. Here there are many choices, as considered below.

Those are the main lines and recommendations chosen for this book. However, we will start with a line that is not considered a main line at all, but is still seen in tournament practice from time to time. Here Black rejects 8 . . . bS , the move that changes this from the Lasker-Larsen variation to the Sveshnikov.

Game 18 Yemelin - Kharlov St Petersburg 1 998

1 .e4 c5 2.�c3 �c6 3.�ge2 �f6 4.d4 cxd4 5.�xd4 e5 6.�db5 d6 7.J.g5 a6 8.�a3 J.e6

This is the Lasker/Larsen/Pelikanvariation - or at least these are some of its many names. This line had disappeared from practice some time ago, but Kharlov has tried to revive it, and thereby challenged his strong opponents to show why it is untenable. 9.�c4 gc8

9 . . . i.e7 10 .i.xf6± . 10.�d5 ixd5 1 1.J.xf6!

This point is what gives White the advantage. Now the light squares are too weak to be justified by a fleeting initiative.

92 The Sveshnikov

1 l • • • gxf6 12.Y!Yxd5

12 • • • c!Lld4 Maybe the best of Black's rather sad list

of choices. 12 . . . ctJ b4 leads to a position with

opposite coloured bishops where Black can only hope for a draw, but will certainly lose more than one game in ten, which is where the 45% score which is the minimum any acceptable Black opening can offer a player slips out of sight. 1 3 .Y!Yd2 d5 14.exd5 ctJxc2t ( 14 .. .'�'xd5 1 5 .'lWxd5 ctJxd5 16.0-0-0;1; is simply more comfortable for White. Black can often reach this kind of endgame right from the opening. 16 . . . ctJ b4 17.c3 ctJc6 I s .id3 ic5 1 9 .ie4 Eic7 20. ctJd6t ixd6 2 1 . Eixd6 r::ll e7 22. Eid3 b5 23. r::ll b l h6 24. Eihdl ± Kurnosov - Zhang Pengxiang, Cappelle la Grande 2002.) 1 5 .'�xc2 ib4t 1 6. r::ll dl and now:

a) 1 6 . . . b5 17.'lWe4! The logical thing is to get out of the pin. 17 . . . bxc4 I S . ixc4

'lWb6 19 . r::ll e2!? {This move makes most sense to me. The king will find safety on the kingside and the rooks get connected. 19 . Eic1 id6 ( 1 9 . . . ic5 was played in Matulovic - Simic, Yoguslavia 1 9S0. Now strongest was 20. 'lWf5 r::ll e7 2 1 .ib3±) 20. Eic2 r::ll e7 2 1 . r::ll e2± Schandorff - Morovic Fernandez, Copenhagen 1 9S2.) 1 9 . . . id6 ( 1 9 . . . ic5 20.b3 ixf2 is possible, and probably best answered by 2 1 .ixa6± when White's advantage is beyond question. However the tempting 2 1 . 'IW f5? did not work because of 2 1 . . . Eixc4! 22.bxc4 'lWb2t 23. r::llf3 id4 24. 'IW cst r::ll e7 25. 'IW c7t r::llffi 26. 'IW cst and White has no more than a draw, �Y2 Filipenko - Sveshnikov, USSR 1 9S0.) 20.b3 (20.ib3 r::ll e7 2 1 . Eihc1 Eic5 22. Eixc5 'lWxc5 23. Eidl 'lWb5t 24. r::ll f3 'lWc5 25 .g3± Bhend - Svedenborg, Lugano (ol) 1 965.) 20 . . . EigS 2 1 . Eiac1 r::ll e7 22.g4 h5 23 .h3 hxg4 24.hxg4 Eig5 25 .id3 EicgS 26. Eic6 'lWbS 27. Eih4+Yastreb - Moskovets, Alushta 2002. White is winning here. Black has no counterplay and is simply a pawn down for nothing.

b) 16 . . J Wxd5t 17. r::ll c1 Black cannot regain his piece as is seen in the following brilliant example. 17 . . . b5 ( 17 . . . 0-0 IS .a3 and White has won this position in several games, one of them being Anka - Tomcsanyi, Hungary 1 995) I S .a3 ie7 19 .ctJd6t r::lld7 20. ctJxcS Eixcs 2 1 .ixb5t 'lWxb5 22. Eidl t id6 23. Eixd6t r::llxd6 24. 'lWxcs 'lWfl t 25. r::ll c2 'lWxa1 26. 'lWxa6t+Berndt - Thiel, Germany 1 995.


Another attempt is: 12 . . . b5 1 3 . tL'l e3

Now Black has tried: a) 1 3 . . . .th6? does not work because

of the following tactical solution. 14. tL'lf5! tL'lb4 1 5 .tL'lxd6t @d7 ( 1 5 . . . @e7 16.Wfxf7t @xd6 17J!d1 t+-) 16.Wfxf7t Only one of several winning moves. 16 . . . @c6 ( 16 . . . @xd6 17.Eid1 t+-) 17.Wfb7t @c5 (17 . . . @xd6 18 . Eid1 t+-) 1 8 .Eid1 tL'lxc2t 19 .@e2 tL'ld4t 20. Eixd4! exd4 21 .Wfd5t @b6 (2 l . . . @b4 22.Wfb3t+-) 22. tL'lxc8t Wfxc8 23.Wfxd4t @a5 24. @f3 and White went on to win in Smagin - Kharlov, Cheliabinsk 199 1 .

13 . . . tL'le7 14.Wfd3 ( 14.Wfb7 Wfa5t 1 5 .c3 Eic7 16.b4 Eixb7 17.bxa5 .th6 18 .tL'lg4 .tg7 19 .a4± Murey - Jamieson, Luzern 1 982.) 14 . . . .th6 (14 . . . Eic6 1 5 . .te2 h5 16.0-0 .th6 17. tL'ld5 tL'lxd5 18 .Wfxd5 Eixc2 19 . .txh5 ± Smirnov - Pilavov, Alushta 200 1 .) 1 5 . .te2 .txe3 16.fxe3;!; Korneev - Hernandez Montalvo, Padron 2002.

b) 12 . . . f5!? is the newest attempt in this position. After 1 3 .0-0-0 b5 14. tL'l e3 .th6 1 5 . @b 1 White is simply better. 1 5 . . . .txe3 1 6.fxe3 fxe4 17.Wfxe4 Wfc7 1 8 .g3! A nice move that exploits the weak structure. 1 8 . . . tL'l e7 1 9 . .th3 Eid8 20.Wfg4 Eig8 2 1 .V9h5 ± Al Sayed - Sveshnikov, Dubai 2004. 13 . .id3

1 3 .0-0-0! ? with the idea of Eixd4! has also been played, but the text move simply assures the edge without any problems. 13 ... �e7 14.�a5 l3xc4?!

This does not work tactically for many reasons.

14 . . . d5 does not promise Black any happiness . 1 5 . tL'l b6 Wfc5 16.Wfxc5 Eixc5 17.c3! As so often in this line Black's sick structure is so important that White only focuses on containment of the black forces. 17 . . . dxe4 1 8 . .txe4 tL'l c6 1 9.0-O-0± .th6t 20. @b1 Eib5 2 1 . tL'l c4 0-0 22. tL'l d6 Eib6 23. tL'l f5 1-0 Balinov Hausrath, Budapest 1 999.

14 . . . Eic5 1 5 .Wfd2 V9c7 1 6.c3 tL'l e6 17. tL'l e3 .th6 1 8 .0-0;!; Varavin - Kharlov, Elista 1 994.

15.hc4 �xc2t 16.i>e2 �xal 17J'�cl!!±

Was this direct assault on the king something Kharlov had overlooked in his home analysis? My computer still has problems finding it, 6 years of technological advance after the game. Even after this move it takes time for the machine to see that something is terribly wrong.

94 The Sveshnikov

17 •• .J.h6 What else? 17 . . . f5 I S .exf5 vggS is

the best option for Black according to the computer. But with some human assistance, it is possible for the machine to find the following nice winning attack. 19 .�bSt @e7 20.ygc7t @f6 2 1 . vgdSt @xf5 22.id3t @g4 23.vgd7t f5 (23 . . . @hS 24.vgxf7t @h6 2S .yge6t @hS 26. Ek4+-) And now a move that takes only a few seconds for the machine to find.

24.h4!! 'it>xh4 2S.Elhl t 'it>g4 26.ixf5t ygxf5 27.f3t 'it>gS 2S.ElhSt 'it>xhS 29.ygxf5t 'it>h6 30.ygf6t 'it>hS 3 1 .g4 mate. To a human this line might seem rather fanciful, but to the machine it is pure logic. 18 . .txf7t! 'it>xf7 19Jk7 Eld8?!

Slightly better was 19 . . . EleS 20Jhe7t �xe7 2 1 . ygc3 �f8 22.b4 �d7 23.vgxal dS 24.exdS �xdS ± - Yemelin. 20.YGd5t 'it>f8 2 U�xe7 'it>xe7 22.�xb7t

gd7 23.�e8+- d5 24.�g8! if8 25.exd5 'it>e8 26.�e6t fie7 27.�xa6 gxd5 28.�c6t :ad7 29.�c1 �b3 30.axb3 'it>f8 31 .b4 gd4 32.�e8t 'it>f7 33.b5 gb4 34.b3 h5 35.g3 fid6 36.�d7t fie7 37.�c6 id8 38.�d5t 'it>e8 39.�d6 gd4 40.�e6t 1-0

From this we will move to the main line, and only after going through it in two games will we pay attention to the sidelines.

In this first game with the main line I have given the most important options until 2o .ixf5!? , which I do not consider critical at all. I just liked the game and would rather include a few more moves than have one game going on for S pages or so.

Game 19 Zeleic - Zelenika Pula 1 999

1 .e4 e5 2.�f3 �e6 3.d4 cxd4 4.�xd4 �f6 5.�c3 e5 6.�db5 d6 7..tg5 a6 8.�a3 b5 9.ixf6 gxf6 10.�d5 f5

l o . . . ig7 l 1 .c3 f5 is nothing but a transposition of moves after 1 2.exf5. Note that Black cannot play his standard idea of 1 1 . . .ttl e7?! here, as White gets the advantage with 1 2. ttl xe7 vgxe7 1 3 .ttlc2± when Black cannot play his regular push on the kingside because of a simple double threat: 13 . . . f5? ! 14.exf5 ixf5? I S .vgf3+-. 1 l .e3


1 l ••• J.g7 1 1 . . .fxe4? is known to be a mistake as

Black is in trouble after 12 .i.xb5 axb5 1 3 .ttJxb5 The main points are: 1 3 .. J '!a4 ( l 3 .. J !a7 14 . ttJxa7 ttJxa7 1 5 .'1Wa4t+-) 14. ttJ bc7t rj;>d7 1 5 .'�·g4t+-. 12.exf5 J.xf5 13.ttJc2 0-0 14.ttJce3

14 . . . i.g6!? is maybe a move for the future. 1 5 .a4!? This move apparently ignores the fragile placement of the bishop on g6. However the usual ( l 5 .h4 is not that convincing. 1 5 . . . i.e4! is Leko's idea, after which Black apparently is doing OK.) 1 5 . . . b4 ( l 5 . . . bxa4 is probably best answered with 1 6.h4! i.e4 17 .'!MI' xa4. The point behind 1 5 .a4, at least when I played it. 17 . . . i.xd5 I S .ttJxd5 ttJ e7 1 9 .i.c4;!;) 1 6. ttJ xb4! Obviously White does not want to accept a strong knight on d4. 1 6 . . . ttJxb4 17.cxb4 e4?! ( l 7 . . . f5! ? i s the way forward for Black, but after I S .i.c4 rj;>hS 19 .i.d5! I prefer White.) I S.'!MI'd2! f5 1 9 .ttJd5 rj;>hS 20.i.e2 i.f7 2 1 .0-0± and Black has no other path

forward than to enter a horrible position with opposite coloured bishops. Baklan -Lobron, Germany 200 1 . 14 ••• J.e6 15.J.d3 f5 16.0-0

1 6.i.c2 was for some time considered the way to play this line, but Black eventually came up with a forced draw with 1 6 . . .f4 17.'1Wh5 gf7 I S .i.xh7t rj;>f8 19 .i.f5 1WeS! ( l 9 . . . gxf5? 20. ttJxf5 i.xd5 2 1 . gdl !+- Arnason - Birnboim, BeerSheva 1 9S7. The draw after 1 9 .. .'!MI'eS was actually given by Arnason, but it took some time before it was played in tournament games.) 20.i.xe6 �xe6 2 1 . � g4 1Wh6! 22. ttJ f5 (22. ttJ c2?! e4, the same goes for 22.0-0?! e4!) 22 . . . 1W e6 23. ttJfe3= .

16 ••• l3a7 This move seems logical and has been

played many times at the top level. However it is far from the only move.

16 . . . gbS 17.'!MI'h5 1Wd7 I S . gadl rj;>hS 19 .i.c2 b4°o was played in David - Manor,

96 The Sveshnikov

Bikurei Haitim 1997. However this is not really relevant. White gets the advantage after 17. i.c2! when the move g b8 has done very little for Black, while White has ideas of WihS and i.b3, both with advantage for White. Thanks to Thomas Luther for this advice.

16 . . . tZl e7? loses in a very famous way to 17. tZlxe7t Wixe7 18 .i.xf5! i.xf5 19 .tZlxf5 gxf5 20.WidSt+-.

16 ... I!f h8!? has been played several times, and is bound to become more popular if the attack on the 16 . . . Ra7 line by De Firmian and Hector continues to be successful. However White seems to be able to create real problems for Black in this line as well, though the last word is far, far away (no, not in that sci-fi movie!) . 17. WihS Now Black has two choices, neither of them fully satisfactory.

17 . . . i.f7 18 .Wih3 e4 19 .i.c2 tZleS. This was the way Illescas Cordoba played with Black against Judit Polgar. Now Polgar blundered with 20.f3?! after which Black had a strong manoeuvre in WigSt followed by Wi d2 with good play. Instead White should play 20.gfd1 i.xdS 2 1 . gxdS Wif6 22. gad1 gad8 23.f4!. Probably preparation from Topalov and his coach. After this there are many ways for White to create real problems for Black. 23 . . . exf3 24. tZlxf5 fxg2 2S. Wixgl (an interesting alternative was 2S .gSd2!? gg8 26.gxg2 tZlg6 27. gfl t) 2S . . . gg8 26. I!fh1 (It seems that it was possible to play 26. gxd6!? gxd6 27. gxd6 Wif7 28. tZlg3 tZlc4 29.Wie4

i.eS 30. gh6 gg7 3 1 . l!fg2t when White has good chances.) 26 . . . i.f8 27.Wih3 tZlg6 28. tZl d4 (28. gSd4!? is very good for White according to Fritz.) 28 . . . ge8 29. ghS gg7 30. gg1 Wif7 3 1 .WifS Wib7t 32.Wif3 Wixf3t 33. tZlxf3 ge6 34. tZld4 gf6 3S . tZlfS gc7 36. tZle3 tZlf4 37. gh4 gcf7 38.i.b3 gg7 39. ghg4 gxg4 �-� Topalov - Leko, Monte Carlo 2003.

17 .. . e4 1 8 .i.c2 tZle7 19 . 9ad1 gc8 20.f3 i.f7 21 .Wih3 tZlxdS 22. tZlxdS b4 23.fxe4 bxc3 24.bxc3 i.xdS 2S.exdS gxc3 26.i.d3 i.d4t 27. l!fh1 WigS 28.gxfS (28.g3!? gf6! 29.Wigl as 30.Wie2! (30.i.xf5?! gxf5 3 1 . gxd4 gel 32. gdd1 gxd1 33. gxd1 Wie3!=) 30 . . . i.eS 3 1 . gbH) 28 . . . gxfS 29.WixfS Wixf5 30.i.xf5 i.e3 3 1 .g3 gel �-Y2. Anand - Topalov, Sofia 2004.

16 . . . e4 is discussed in Game No. 21 below. 17.a4!

17.WihS gaf7 has been played many times. The conclusion is that Black is doing


fine, so there is no real reason to go further down this dead-end road.

I think that 17.a4 was actually an idea of co-author Golubev. 17 ... �e7

This move does not seem to offer Black enough compensation for the pawn. However the alternatives are also not recommendable. 17 . . . bxa4 1 8Jha4 a5 ;!; Rogozenko. 19.�h5 e4 20. �f4 if7 21 .ic4 ixc4 22. ltJxc4 �d7 23J'!:fal ± 1-0 Kolcak-Kucinskas, Email 2002.

17 . . . b4 also does not seem to be very promising. 1 8 .cxb4 ( 18 .�h5 e4 19.1tJf4 if7 20.ic4;!; - Golubev) 1 8 . . . f4 ( 1 8 . . . ltJd4!? i s unclear according to Golubev. However this seems a bit superficial. 1 8 . . . e4 19 .tt:lf4 if7 20.ic4± Pedersen - Aagaard, Aarhus 1999 is my own sad experience with this position.) 19 .�h5 �ff7 ( 19 . . . ih8 20.ifS �xf5 2 1 . ltJxf5 ixd5 22J'!:fd1 ltJxb4 23.�ac1 if6 24.ltJxd6!, Areshchenko - Holmsten, Cappelle la Grande 2003) 20.ixh7t cj;lfB 21 .ifS �xf5 22. ltJxf5 ixd5 23.b5;;;; Vallejo Pons - Shirov, France 2002.

17 .. .f4 18 .�h5 ih8 ( 18 . . . �ff7 19.ixh7t!? cj;lfB 20.ifS!± - Golubev) 19.ixh7t �xh7 20.�g6t+-.

17 . . . e4 18 .ltJf4 if7 19.axb5 ±. 18.�xe7t

White has two alternatives, one cautious, and one wild and hot headed:

18 .ic2!? ltJxd5 1 9. 1tJxd5 bxa4 20. �xa4;!; is certainly possible. Now after 20 . . . cj;lh8 2 1 .�d2 a5 22. �dl �b7 23.b3

id7 24.�a2 �b5 25 .c4 �c5 26. ltJc3? Black was OK in Wedberg - Von Bahr, Stockholm 1999, but instead White could have played 26. ltJe3 �f6 27.�da1 ±. 1 8 .axb5!? ltJxd5 (18 . . .f4?! 19 .1tJxe7t �xe7 20. �h5+- Golubev-Horvath, Scuol 200 1) 19 .ic4! (inferior is 1 9.ixf5?! �xf5! 20. ltJxf5 a5 !! 2 1 .c4 ltJf4 22. ltJxd6 ifB 23.c5 (23 .ltJe4 �h4+) 23 . . . �d7!, Nijboer) 19 . . . ltJf4 20.ixe6t ltJxe6 2 1 .�d5°o Topalov - Leko, Monte Carlo 2004. 18 .. J�xe7

1 8 . . . �xe7? 19.axb5 axb5 20.�xa7 �xa7 2 1 .ixb5. Why not? In Navara -Hansen, Bled 2002 White also achieved a good game with 2 1 .ic2, but this is more convincing. 2 1 . . .�c5 22.ia4 f4 (22 . . . d5 23.b4 �d6 24.ib3±) 23.ib3!±. 19.axb5 axb5

2o.ix£5!? In the next game we shall look at

the critical 20.ixb5, but here we shall investigate a few side options. 20. �a6 d5

98 The Sveshnikov

does not give White anything besides the option to transpose again with 2 1 .ixb5! . The following examples should make this clear:

21 .�h5 e4 22.ixb5 f4 23Jhe6 l"lxe6 24.�xd5 �xd5 25. l2hd5 l"lb8 26.!c4 @h8 27. lt'lxf4 l"lee8 28.b3 ixc3= Nilsson - Binelli, Corr. 1998.

2 1 . l"lxe6!? l"lxe6 22. lt'lxf5 e4?! 23. lt'lxg7 @xg7 24.ixb5;1; Zelcic-Sermek, Belisce 1999.

After21 .lt'lc2 .ic8 22. l"la8 �d723. lt'lb4 e4 24.ie2 ib7 25. l"la5 d4 26.cxd4 Black has no problems. Y2-lh, Anand - Kasparov, Moscow 2004. Though improvements on White's play have been suggested, they do not include any chance of an advantage.

20.ic2 @h8! also fails to impress. 2 1 . l"la6 (2 1 .ib3 f4! 22.ixe6 l"lxe6 23. lt'ld5 f3!,,) 2 l . . .e4 22. lt'ld5 (22. l"lxd6 �b8!?�) 22 . . . l"lee8 + Rogozenko. 20 •• .lhf5

20 . . . !xf5? does not work because of 2 1 .lt'lxf5 l"lxf5 22.�d5t l"lef7 23. l"la8+and White wins the queen. 21.tiJxf5 .bf5 22.Y1Y8 Y1Yd7

Black should defend with the bishop close to his king, as the pin is very uncomfortable after 22 . . . ic8? 23.�d5t @h8 24.l"la8± . 23Jl:a8t if8 24J�fal gf7 25.Y1Ye3!?

With this move White simply gets out of the open file. The grandmaster might have known that 25. l"lb8!? ie6 (25 . . . ig6 26. �d5 �f5 27.f3;1; Rogozenko.) 26. � g3t probably leads to draw after

26 . . . l"lg7 27.�f3 l"lf7 28.�g3t= as shown by Rogozenko.

25. l"l 1 a7?! is a very bad idea. 25 . . . �xa7 26. l"lxa7 l"lxa7 27.�d5t @g7 28.f4 @f6 29.fxe5t dxe5 + It is difficult to evaluate how much worse White is in this endgame, but the main point is that he went on to lose it in the game David - Yakovich, St Vincent 2000. 25 . . • Y1Ye6 26.l1b8 Y1Yd5 27.Y1Yg5t igG 28.M Y1Yc5 29.Y1Yg3

Obviously Black is not seriously worse here. However it is of little theoretical importance. 29 ••. e4 30.ga5 l3f5 31 .b4 Y1Yd5?

Black should not insist on protecting the b-pawn with the queen. Now White had a very strong continuation. 32.Y1Ye3?

32. l"laxb5! Of course either rook can capture here. 32 . . . �xb5 33. l"lxb5 l"lxb5 34.c4! l"lxb4 35 .h5 ±. 32 •.• l3h5 33.Y1Yf4 gf5 34.Y1Yc1?!

34.�e3=. 34 •.• ih5?!

Very strong was 34 . . . �d3!+ . 35.l1al?!

35.l"laxb5 �xb5 36. l"lxb5 l"lxb5 37.c4;1; was still good for White. 35 .•. Y1Ye6 36.Y1Ye3 Y1YgG 37.�f1?

37. @h2 �g4 38.�g3 with some chances for a draw. 37 •• .i8?!

37 . . . � g4! was very strong here. 38.g3?!

38.gxf3 was stronger, as Black cannot


win. 38 . . . exf3 39.Wfd4 Wfg2t 40. 'it>el !'le5t 4 l .'it>dl Wfflt 42. 'it>c2 Wfxal 43.Wfg4t 'it>f7 44.Wfxf3t 'it>e6 45.Wfg4t 'it>f7 46. Wff3t=. 38 ... YlYg4 39.�el E:d5??

Deep in time trouble Black goes in for the kill, but leaves his own king without a defence. White now wins in style.

Best was improving the king's safety with 39 . . . 'it>g7! 40.Wfd4t 'it>g6 4 l . !'lxf8 !'lxf8 42. !'la6 !'lf6 43.!'lxd6+.

40.gxf8t! Now White wins.

40 ... �xf8 41.YlYh6t �f7 42.ga7t �e8 43.E:aSt �f7 44.E:f8t �e7 45.YlYf6t �d7 46.YlYd8t �c6 47.YlYaSt �c7 48.E:f7t 1-0

In the next game we shall see my main recommendation. The Swedish grandmaster delivers a great concept in the opening to shake the Norwegian wonderboy. The line is based on an idea of Nick De Firmian.

Game 20 Hector -Carlsen Malmo 2004

1 .e4 c5 2.�f3 �c6 3.d4 cxd4 4.�xd4 �f6 5.�c3 e5 6.�db5 d6 7.1g5 a6 8.�a3 b5 9 . .bf6 gxf6 10.�d5 f5 1 1 .c3 i.g7 12.exf5 ixf5 13.�c2 0-0 14.�ce3 .te6 15..td3 f5 16.0-0 ga7 17.a4 �e7 18.�xe7t E:xe7 19.axb5 axb5 20.ixb5

This should be the real test of Black's pawn sacrifice. However the pawn cannot be accepted without a valid idea to follow. And that idea should probably be to triple the heavy pieces on the a-file and invade Black's position on the 8th rank to force exchanges. 20 ... d5

21 .ga6! White is preparing the best formation

of the heavy pieces: Ra6, Qa4 and Ral . Furthermore, the rook is well placed for several tactics involving Rxe6, and

1 00 The Sveshnikov

this is very unpleasant for Black. Still, it is worthwhile to take a look at the alternatives provided by this position:

a) 2 1 . '%lI a4!? is the most promising of the alternatives, even though I do not think it should be really dangerous for Black. 2 1 . . .f4 22. tt:lc2 E:ef7 (22 . . . f3?! Generally Black should not make such rash decisions concerning the attack. Now he has lost a lot of flexibility and it is much easier for White to defend the kingside, and thereby also to operate freely on the queens ide. 23 .g3 E:b7 24. tt:l b4 '%lIg5 25 .E:fd1 ± Here I have great doubt about the compensation for the pawn. Yefremov - Gilbert, Corr. 1 995.) 23. E:fd1 '%lIg5 24.�c6 e4 25. '%lIb5 e3 26.f3 (Maybe it is possible to improve here with 26.�xd5!? which of course seems very greedy, but it is not so easy to refute. Black has no easy way to crash through on the kingside. A possible line could be something like 26 . . . exf2t 27. �xf2 f3 2S .g3 '%lIh5 29.h4 �e5 30. E:g1 ! �hS 3 1 . tt:l e3;!;) 26 . . . �hS 27. E:a5 '%lIh4 2S.'%lIe2 E:bS 29. E:b5 E:xb5 30.�xb5 E:b7 l!2-l!2 Szczepankiewicz -Kruse, Email 2000.

b) 2 1 . �c6 is not dangerous for Black. After 2 1 . . .d4 22. E:aS '%lId6 23. E:a6 he can play 23 . . . QdS when White has nothing more than a repetition of moves, or even go for a better game with 23 . . . �h8 24.cxd4 exd4 25. tLlc2 '%lIc5 26.b4 '%lIc3 27.�b5 �b3 2S.�d3 E:c7 29. E:a3 '%lIb2 which was altogether bad news for White in Smirnov - Sitnikov, Russia 200 1 .

c) 2 1 . tt:l c2 is also feeble. The best White can hope for is probably the drawish line chosen by Kramnik. Jelen's idea seems to be risk free for Black, but certainly dangerous for White. 2 1 . . . E:b7 22.�c6 E:b6!? (22 . . . E:xb2 23. E:aS '%lId6 24. E:a6 '%lIc5 25. tt:lb4 '%lIxc3 26.�xd5 '%lIxb4 27.�xe6t 1 12-1 12. Svidler - Kramnik, Wijk aan Zee 1999) 23. tt:l b4 d4 24. E:a7 e4� Skyrte - Jelen, Budapest 1 999. 21 . .. i>h8!?

Rogozenko writes that this is clearly the best move. He has no faith in 2 1 . . . f4 because of the obvious exchange sacrifice 22. E:xe6! E:xe6 23.'%lIxd5 (To me it actually make sense to investigate 23. tLlxd5!? as well. The best is probably the computer's suggestion 23 . . . �hS (23 . . . f3 24.g3 '%lIbS 25.c4;!;) 24. E:e1 e4oo) 23 . . . '%lIxd5 24. tt:lxd5 Now the best option is 24 . . . �hS! (24 . . . f3?! 25 .E:e1 fxg2 26.�d3 E:dS 27.�e4± van Kempen - Arduman, Email 2000.) 25 .f3;!; Rogozenko. Staudler - Hohm, Corr. 1 999. Also worth looking at is 25 .b4!? e4 26.c4 e3 27. E:e1 �d4 2S. �f1 and White looks better here, doesn't he?

Anyway, the main conclusion about 2 1 . . .f4 is that it is hard for Black to prove complete equality. It would be sad for those defending the Sveshnikov to have to resort to this kind of position. If this really were the case the popularity of the opening would vanish overnight. However, it is also not pleasant to be Black in the main game, so maybe it is

Experts vs. the Sicilian 1 0 1

necessary to look for improvements for Black earlier, maybe as early as move 1 6?

22.V!Ya4! An idea by the American grandmaster

living in Denmark, Nick De Firmian. No other move seems to give White a position he can really feel good about playing.

22. lLl c2?! llNb8 23. V!Y e2 f4 24.f3 i.c8 has given Black a good game on several occasions. It is a dead end we do not have to go down.

22.llNh5?! f4 23. lLlg4 llNb8 is also nothing. Hodova - Priborsky, Plzen 2003.

22.f4?! is, according to Rogozenko, "the most logical move, even if Black wins an exchange by force after that." I do not understand the logic that encourages advancing pawns in front of your own king and losing material in the process. This is at least not the classical way to view such situations. Rogozenko of course has good reasons for his assessment ( .. .f4

was coming) , however I think the balance is tipping in the wrong direction. Now after 22 . . . d4 23.cxd4 exf4 24J'!:xf4 i.c8 25J:!:a3 i.h6 26. Ei:f3 f4 27. lLlc2 i.g4 the game was very complicated in Hector -Rogozenko, Gothenburg 2004. However, I do not think that the first player should be very happy about the outcome of the opening. Eventually Black managed to win this game, though this was hardly the only possible outcome. 22 ... f4

Obviously this was what Hector feared in his game above against Rogozenko. However, it is clearly the lesser of two evils, as Black has no easy way to break through White's defences. 23.c!Dc2 gg8

This is the first new move of the game. In the stem game of 22.llN a4 Black chose a less natural, but still somewhat logical move 23 . . . i.f5 which tries to provoke the white knight into occupying an unwanted square. However, Black also loses important time and White should be able to gain an advantage. 24. lLl b4 (24. Ei:dl ! ? could also turn out to be a good move. It is not clear that opposite coloured bishops will be an advantage for Black in the coming double-edged position. However De Firmian's choice looks good, so we will focus on that here.) 24 . . . Ei:ef7 25 .Ei:dl V!Yg5 26. lLlxd5 i.g4 (26 . . . e4!? 27.i.f1 e3 28. 'it>hl ±) And now I would suggest following Fritz 8 which gives 27. Ei:d3! (And not 27. Ei:d2?! i.f3 28.i.f1 e4� De

1 02 The Sveshnikov

Firmian - Svensson, Gothenburg 2004, which eventually ended in a draw, where White was the one defending.) 27 . . . !f5 28J i:d2 f3 29. ttJ e3± The position is still very complicated, but Black's attack is still restrained and it seems likely that White will be able to benefit from his extra pawns.

23 . . . �c8 24J �a8 �e6 25 .�al �h6 26.'1Wb4 VNh4 27.h3 �g6 28.�d7! �xd7 29 . �xfllt �xfll 30.'�·xfllt �g8 3 1 . VNd6 �xg2t! with a draw by perpetual check was the correspondence game TeichmannMarotta, 2003. But afrer 26. ttJe l ! ;!; I prefer the White pieces.

24J�al!� Once it has been established how

White should organise his pieces this move becomes very logical. White needs to bring the rook into play for many reasons, but the urgency is because he needs to play ttJe l very soon to protect g2, the sofr spot in his position.

The computers love for 24. ttJ b4?! is completely unjustified. The knight has to go to e l to protect the king. 24 . . . �fll !� gives Black good play. Now White should take the draw promised to him in the tactical lines, or everything might soon be very bad. 25 . ttJ c6 (25 .g3? fxg3 26.hxg3 �eg7! ! 27. �xe6? VNh4-+) 25 . . . �xg2t 26. i>hl �g4!! (26 . . . VNe8 27.�e2 �eg7 28. ttJxe5±) 27. i>xg2 (27. ttJxd8 !f3 28.h3!=) 27 . . . �g7 28. ttJxd8 (28.f3?? �h3t! ! 29. i>£2 11*lh4t 30. i>e2 �xf1 t 3 1 . i>xfl VNxh2 32. i>e1 �g2 33.�f1 �d2-+) 28 . . . �dl t 29. i>h3 �g4t=.

24. �dl ! ? is the only natural alternative to the text move, and could be used as a surprise move against someone who thinks the position afrer 24.Ral is playable. However, I have a feeling that well informed Black players will tend to avoid this in the future, once it has been tested a few times at the top level. 24 . . .i.c8

Maybe a new idea can be conceived here for Black. However I cannot see that Black gains anything with 24 . . . f3 25 .g3 �c8 26. �a8± . 25J�a8 J.m

Black is getting ready to create threats against g2. However they are not strong enough and they come too late. 26.J.fI

This seems to favour White, but it all becomes very complicated now. Strong was the direct 26.�c6! and now it seems to be difficult for Black to defend against


threats like 'i;lfb5 and !l:dl . The important thing for White is that iLl el will protect the kingside almost single-handed. 26 . . . !l:c7 (26 . . . !l:eg7 27. iLle l 26 . . . !l:xg2t 27. 'it>xg2 !l:g7t 28 .'it>hl ig4 29. iLle l +-) 27. !l:dl 'i;lfg5 2S. iLle l ± . 26 .•• Eieg7 27.YlYc6 �k7

The only move. 28.YlYb5 .ic5?!

The young Norwegian is desperately looking for counterplay, however he would have been better off asking White to prove his advantage after 2S . . . !l:cg7!? 29. !l:dl 'i;lfc7! This is not so clear, as after 30. !l:xd5 id7 Black will win the exchange with . . . Bc6 and the endgame is not that bad for him. Maybe 29.c4!? would be stronger?

29.M hflt!? Carlsen goes into these tactics searching

for some action, as he is likely just to be run over after 29 . . . ifll 30.c4± . It is always possible to dismiss such desperate

measures after the game, especially armed with thorough computer analysis. However from a practical point of view this was obviously the right decision, as Black gained a lot of ground in the remaining part of the game, and should have played on when it finished. 30.c;t.>xfl gxc3 31 .�e1

Not the only choice, but certainly one that makes sense. 31 ...f3!?

Again complicating matters. Black opens files and gains time at the cost of yet another pawn. 32.�xf3?

Probably kingsafetywas more important than anything else in this position. After 32. 'it>gl ! it is not easy to see how Black would be able to attack White's king. 32 . . . 'i;lfh4 (32 . . . e4 33 . !l:Sa2 and what now?) 33. 'it>hl Here I cannot work out a method to create a successful attack against White's king. 33 . . . 'i;lff2 (33 . . . !l:e3 34. !l:xc8 !l:xcS 35 .iLlxf3 'i;lfe4 36.'i;lfd7 !l:fll 37. !l:a7 'i;lfg6 3S.'i;lfxd5+-) 34. !l:Sa2 'i;lfd4 35 .'i;lfa4 ih3 36.'i;lfdl +-. All of this is of course still very complicated, and all the conclusions should be seen as temporary. Still, I feel that White should be successful. 32 ... e4 33.�el YlYf6t 34.c;t.>gl �U8 35J��8a2

35 . iLl f3? is the computer's favourite move for some time. A human would hardly consider this, and after 35 . . . exf3 36. 'i;lfxd5 !l:c2 !+ it all turns out to be

1 04 The Sveshnikov

horrible, as it should be. However, as we shall see, the idea is not at all stupid.

35 ••• ga3? This was the apparently brilliant idea

conceived by Carlsen. The rook cannot be accepted of course, because of '%Yf2t. However, White still has a chance to make his extra piece count, by returning it if nothing else. Therefore a simpler method of play, achieving instant repetition of moves, was preferable.

35 . . . '%Yd4t 36. c.:t>hl id7 37.'%Ya6 icB 3B .'%Ye2 ig4=.

Note that 35 . . . id7 36.�e2 ig4? would not work because of 37.'%Yf2!± . However, 36 . . . '%Yd4t would still draw. 36.Y!Ye2?

This should probably have been punished by a strong tactical resource. However with little time on the clock it is hard to find the right path through such a tactical maze.

36.g3 ia6 37.'%Yxa6 �d4t 3B . c.:t>hl

i'!xa6 39.ixa6 �xb4 and White i s maybe on the way to being worse.

36. liJ£3! was the best move. After 36 . . . i'!xa2 37. i'!xa2 ex£3 3B . i'!f2! Black still has not solved his problems. Actually there is no way for him to save both the d5-pawn and the £3-pawn, so a sad defensive task awaits him in a 3 pawns vs. 2 pawns endgame. 36 ••• Y!Yb6t 37.<i!?hl Y!Yf6 38.<i!?gl Y!Yd4t l/2-l/2 A draw was agreed. However in this position it was Black's turn to play on.

39. c.:t>hl could be answered strongly with 39 . . . ig4! and now White seems to be worse:

40.'%Yb5 �'lxa2! (40 . . . '%Yxal 4 1 . i'!xal i'!xal 42.h3 i'!xe l 43.hxg4 i'!exfl t 44. c.:t>h2 i'!dl 45 .�d7 and White should be able to deliver perpetual check) 4 1 . i'!xa2 �dl 42.h3 �xe1 43. c.:t>gl (43 . c.:t>h2 �xfl 44. '%Yxfl i'!xfl 45.hxg4 d4-+) 43 . . . id7 44.'%Ye2 '%Yxb4-+

40. liJ £3 i'!ax£3! 4 1 . i'!aB i'!xaB 42. i'!xaBt c.:t>g7 and White is definitely in trouble. 43 .'%Yel �f2 44.'%YaI t i'!f6 45.ib5 d4+ .

However, Black was short of time and a draw seemed to be a good outcome from such a bad experience in the opening.

In the next game we shall investigate a minor sideline that offers little hope for Black. In fact the analysis goes in the direction of a clear edge for White in the opening, so it is probably not here that the future lies for Black. Still, it is wise for

White to know this line in case i t should come up in a real game. Not everything is easy to find over the board.

Game 21 Rivas Romero - Sarlat Corr. 2000

1 .e4 c5 2.�f3 �c6 3.d4 cxd4 4.�xd4 �f6 5.�c3 e5 6.�dh5 d6 7.�gS a6 8.�a3 h5 9 . .ixf6 gxf6 10.�d5 ig7 l l .c3 f5 12.exf5 hf5 13.�c2 �e6 14.�ce3 0-0 15.id3 f5 16.0-0 e4

One commentator writes that Black has scored well with this approach recently. I do not know where he has been looking, as to me it seems that Black is actually not doing very well in this line. 17.�f4

Obviously 17 .!c2? f4 1 B .!xe4 fxe3 19 .YMh5 exf2t 20. <;!;>h1 �f7+ spells disaster. Bestagno-Kuntz, Cagnes 1 9B9.

17 .•• W 'The alternatives are: 17 . . . !d7!? 1 B .!c2

!e5 1 9 . ct:\ ed5 <;!;>hB 20.'�lh5 �gB 2 1 .8;1; with better play for White in Burnoiu -Veneteanu, Curtea de Arges 2002.

17 . . . 'Wd7 1 B . .ic2 �e5 1 9.tiJxe6 was played in Rogovoi - Nikolaev, St Petersburg 1 99B. However, stronger seems to be 1 9 .tiJed5!? l'%aeB 20.�xe6 l'%xe6 2 1 .a4. 18.J.c2 �e5

1 B . . . YMd7 19 .!b3 �adB 20.!xf7t YMxf7 2 1 .�d5 YMxd5 22. ct:\exd5 ;1; Mamedov - Djafarli, Baku 2002. 19.�fd5

1 9 .93 !xf4 20.gxf4 YMf6 is even according to van WeIy/Cifuentes. 19 ... WgS

1 9 .. .f4? 20. ct:\g4 !g6 2 1 . �e1 e3 22.fxe3 !xc2 23.YMxc2 �g5 24. ct:\xe5 �xe5 25 .<;!;>h1 ±, Kramnik.

1 9 . . . YMh4 20.g3 YMh3 (20 . . . �g5 2 l .f4 ex8 22.�x8 !h5 and now 23.�f2 was played in Palecha - Mikhajlichenko, Evpatoria 200 1 . Instead 23. � g2! ± would have been very strong.) 2 1 .f4 !g7 22. ct:\ c7 (22.!b3 <;!;>hB 23. ct:\c7 !h5 24.�d2± Lantini - Frilli, Arco 1 999.) 22 . . . :1'iadB 23.!b3 d5 24.!xd5 <;!;>hB 25.YMe2 �d7 26.!xf7 �xc7 27.!b3 ct:\d4 2B.cxd4 1-0 Collazo - Bianchi, Email 1 999.

19 . . . !e6 20.f4 ex8 2 1 .�xf3± f4? 22.�e4 �a7 23. ct:\xf4+- Grabarczyk -Rydzik, Zakopane 2000. 20.f4 ex8

1 06 The Sveshnikov

21 .E:xf3?! Probably stronger is 2 1 . Wl xf3 when after

2 1 . . .�h5 22.Wlh3 f4 Anand - Kramnik, Linares 1 995 23. l2k7! is very strong. (23 . ctJ f5 �hS 24. ctJh4 :1'1a7 25. ctJ f3 �xf3 26. Wlxf3 :t Longson - Son, Istanbul 1 995.) 23 . . . fxe3 24. ctJ xaS :1'1xfl t 25 .:1'1xfl ± was played in Karasek - Underwood, e-mail 200 1 . White is clearly better according to Kramnik. The reality is probably that he is completely winning. 21. . . .ih5

van Wely analyses 2 1 . . .f4 and comes to the conclusion that 22.h4! gives White an advantage. 22 . . . WldS 23 .ctJxf4! Wlxh4 24. :1'1h3! Wlxf4 25 .�xh7t �g7 26. ctJ f5t �f6 27. :1'1h4!+-. 2vllJc7 .if4

This does not seem necessary, but Black still should not be worse. 23.Wd5t <j;1h8

23 . . . �f7 24.Wlxc6 �xe3t 25 .�hl +-. 24.E:xf4

This is, of course, the correct bishop to eliminate. 24 ... Wxf4 25.E:el

Also possible was 25 .ctJxf5 !? ctJ e7 26.Wlxd6 Wlxd6 27. ctJxd6 :1'1a7 2S. ctJ e6 :1'1f6=. 25 .. . tLle5

Surelyan improvement over25 . . . :1'1acS?? 26. ctJ e6 Wlh4 27.g3 :1'1gS 2S .ctJxf5+Topalov - van Wely, Wijk aan Zee 1 999. 26.Wd4!

Necessary. White cannot accept the rook, as Black would naturally ignore the knight on as and instead go for the king. 26. ctJxaS? :1'1gS!! 27. Wlxgst (Or 27.g3 :1'1xg3t 2S.hxg3 Wlxg3t 29. �fl (29. ctJ g2 �f3) 29 . . . Wlf4t 30. �gl ctJ f3t-+ with a winning attack.) 27 . . . �xgS 2S.�xf5 Wlh4 29. :1'1fl Wlg5-+ van Wely. 26 ... Wxd4

Now there follows a series of forced moves that leads to an endgame that Black plays quite badly. Of course, we can all have bad days, but it must feel terrible when you play correspondence chess and have so many of them in the same game. 27.cxd4 E:ac8 28.tLle6 E:fe8 29.dxe5 E:xc2 30.tLlxc2 E:xe6 31.tLld4 E:xe5 32.E:xe5 dxe5 33.tLlxf5

This should be a draw, but realistically only White can win. 33 .. . .ig6 34.tLld6 <j;1g7 35.tLl b7 .if! 36.a3 .id5 37.tLlc5 a5 38.tLld7 e4 39.<j;1f2 .ic6 40.tLlc5 <j;1f6 41 .@e3 @e5 42.g4 h6 43.h4 a4 44.g5


44 . . . h5? As I said, Black's play has not been great.

One move that was hard to understand was 43 . . . a4, but that was hardly the only mysterious move. Now he chooses to keep the h-pawn on the board. All endgame experts agree that when you try to defend, you aim for the exchange of pawns. Here Black could probably have held a draw with 44 . . . hxg5 45.hxg5 1iff5 46. lifd4 Iifxg5 (46 . . . e3! ? is perhaps even better.) 47. llJxe4t Iiff4 48. llJ c3 i.d7 49.lifc5 life3 50. llJxb5 lifdl 5 1 . llJc3 Iifc2 and, with only one pawn remaining, White should not be able to win this endgame. 45.g6 mf6 46.fllxe4t mxg6 47.mf4 mf7 48.fllg3 mg6 49.me5

Now Black can no longer save the game. The two weaknesses and the bad bishop seal his fate. 49 .• .i.e8 5o.flle2 !J.f7 51 .fllf4t mg7 52.flle6t mg6 53.flld4 J.c4 54.flle6 .ifl 55.fllf4t mh6 56.mf6 Ac4 57.flle6 .ifl 5s.flld4 1-O

Finally, I will end this chapter with what I believe will be a main line in the Sveshnikov Sicilian in the future. An under-rated player delivers a performance with White of which he can be truly proud. He defeats one of Israel's younger stars in a convincing manner.

Game 22 Jenni - Avrukh Bled (01) 2002

1 .e4 c5 2.flla fllc6 3.d4 cxd4 4.fllxd4 fllf6 5.fll c3 e5 6.flldh5 d6 7.Ag5 a6 s.flla3 h5 9.ixf6 gxf6 lo.flld5 f5 1 l .exf5 hf5 12.c3 Ag7 13.fllc2 .le6!?

The main idea here is to play llJ c6-e7xd5 and liquidate White's pressure. The immediate 1 3 . . . llJe7?! was refuted by Ivanchuk with 14 .i.d3! and White has a clear plus .

I think 1 3 . . . i.e6 is the move Black players will be playing in the future. They

1 08 The Sveshnikov

have not played it a lot for the last few years, but it offers a relatively safe position with good counter chances. I still prefer to be White, but in such a reliable opening as the Sveshnikov it would be too much to hope for to prove an easy advantage for White in every line. That I have come as close as I have is pure luck, and was not something I thought possible in advance. 14.g3!

This was played by Anand a long time ago, but Short made it popular with some good games, including a win against Kramnik. The key idea is that if Black exchanges on d5 then White would like to recapture with the bishop. The plan is that if White ends up with knight vs. bishop, then d5 and f5 are likely to be rather weak squares. 14,..0--0 15.,tg2 a5

This is clearly the main line. Other moves have been played, but I would not recommend anyone outside the world's top 50 players to try to memorise the differences between 1 5 . . . :B:b8 and 1 5 . . . a5 . The main point for White i s that the same set-up is recommended against both options. 16.0--0 f5

This specific line can of course be prepared to a great extent, but many different moves are possible all the time, and it is therefore more important to grasp the essentials of the position. I would not like to claim mastery of the position, but I think that I can give a few hints and

ideas. First of all, I recommend placing the queen in the centre, and out of the way of the rooks. I dislike 17. 'Wh5 as I do not see a fair argument for attacking f5 already, and I dislike 17 . 'W d2 as this should be the place of a rook, not the queen.

17.VNe2! gb8 18.gadl 1 8 . :B:fdl is slightly more popular, but

honestly I prefer this one. Now an early . . . e5-e4 can be met by £2-f3 with much greater strength. 18,. .VNd7

1 8 . . . e4 1 9 .f3!;j; . 19.f4!? e4

1 9 . . . @h8!? 20. lDce3 b4 looks like a sensible plan, but with the help of my computer I managed to find an idea that I believe is unpleasant for Black. 2 1 .fxe5! ? (2 1 . @hl bxc3 22.bxc3 lD e7°o Resika - Sallai, Budapest 2003.) 2 1 . . .!xe5 22. lD c4!;J; . The following line shows how it is possible for Black to end up


with a weaker pawn structure, where the weaknesses are on the light squares, which should be very unpleasant. 22 . . . bxc3 23.bxc3 l':lfd8 24. lLl de3 !xc3 2S . l':lxd6 !xc4 26. lLlxc4 'fie7 27. l':le6 'ficst 28. 'it>h1 lLl d4 29. l':leS 'fib4 30.'fid3± . 20.tilce3 tile7

Black needs to fight for the centre. 20 . . . b4 is strongly met with 2 1 .g4! bxc3 22.bxc3 fxg4 23.!xe4;!; . 21.gd2 tilxd5 22.tilxd5

22 . . . Wf7 22 . . . b4! with the idea of 23.c4 b3

24.a3°o is probably the way to play Black's position. Right here is probably the most difficult place to prove an advantage for White. I have looked at natural moves such as 24 . . . l':lfe8 2S .l':lfd1 'fia7t 26. 'it>h1 'ffcS 27. lLle3 l':lbc8 and now I can only manage to prove a draw for White. 28.!h3! (28 .l':lxd6? !xc4 29. lLlxc4 'ffxc4 30. 'fihS 'fff7!=t because of the weakness of the b2-pawn.) 28 . . . !xc4 29. lLlxc4 'ffxc4

30.!xfS 'fixe2 3 1 . l':lxe2 l':lc2 32. l':lxe4! l':lxe4 33.!xe4 l':lxb2 34. l':lxd6 and White will be able to scrape the draw without any problems. However, this is far from an advantage for White. 23JUdl gfc8 24.tile3!?�

I really think that this is the way forward. 24.g4 fxg4 2S .!xe4 'it>h8 was unclear in Gruenfeld - Sutovsky, Israel 1 996. 24 .. ..ha2 25Jhd6 J.b3 26J�1d2

26 . . . b4? This allows White to enter the 7th

rank, a very careless admission. It was necessary to return the bishop and start to protect the coming weaknesses in advance with 26 . . . !e6 27.ih3! (27.g4?! if8!oo 28. l':lxe6?! 'ffxe6 29.gxf5 'ffh6=t) and now we have:

a) 27 . . . l':le8 28. l':lxe6 'fixe6 29.ixfS 'ffb6 (29 . . . 'fff6?! 30. l':ld7! l':lbd8 (30 . . . h6 3 1 . 'ffhS!+- lLldS will come and Black will not be able to defend his king. Probably it

1 1 0 The Sveshnikov

will transpose.) 3 1 . it:ld5! �fS 32. �h5 h6 33.'!Wg6 l"!xd7 34. it:lf6t �xf6 35 .�xeSt �fS 36. �xd7+-) 30. mg2 (30. mhl l"!bdS 3 1 . l"!xdS l"!xdS 32.�xe4 �e6 33 .'1Wf3 mhSoo) 30 . . . l"!bdS 3 1 . l"!d5 ;!; l"!xd5 32. it:lxd5 �d6 33.'�h5 mfS 34.�xh7± e3? 35 .�5t+-.

b) 27 . . . l"!fS!? Overprotecting 5. 2S. it:l c2 b4 29.cxb4 axb4 30. it:ld4 �c4 3 1 .�f2;!;.

27J�ld7 White is better. 27 .•• Y!YfS 28.cxb4!?

It is possible to understand this move, but not to fully approve of it.

2S .c4!± would quickly have made Black's position fall to pieces, as �a2 is out of play. One critical line is 2S . . . a4!? but analysis suggests that this is not dangerous (2S . . . l"!c5 29.�h3 �a4 30. l"!7d5 and White wins. Probably White was afraid of the a-pawn) 29.�h3! a3 (29 . . . l"!c5 30.'lWh5 'lWeS 3 1 .'lWh4 a3 32. it:lx5+- and there is no defence against the attack.) 30.bxa3 bxa3 3 1 . it:lxf5+-. 28 ... axb4 29.i.h3 i.e6 30J:l7d6 Y!Yf'7 31.l"!xe6 Y!Yxe6 32.ixf5 �kl t 33.�g2 Y!Yb6 34Jld7 gd8

35.Y!Yh5 Y!Yf6? 35 . . . l"!xd7 was the only move. However,

after 36.'lWeSt �fS 37.�e6t! 'lWxe6 3S . Y!Yxe6t l"!f7 39. 'lWxe4+- it is hard to believe that Black would survive. 36.tLld5

One of many winning moves. 36 ... Y!Yxb2t 37.�h3 gxd7 38.Y!Yxh7t

3S .�e6t! mfS 39.�xd7+- was definitely easier. 38 ... �fS 39 • .hd7 gc5

39 .. J'l:hl 40. mg4 l"!xh2 4 1 .'lW5t was still winning. 4O.Y!Yg6 id4 41 .J.e6 1-0


The Classical

-By Peter Wells

1 .e4 c5 2.�f3 d6 3.d4 cxd4 4.�xd4 �f6 5.�c3 �c6

The Classical Variation of the Sicilian has one powerful and enduring appeal. Black's development is faster and healthier than in several other Sicilians, and this tends to afford some protection from the blustery winds of theoretical change. Of course fashion still takes its toll, but a certain stability over time can be expected here. 6.J.g5

The Richter-Rauzer is the most respected antidote, and also combines aggression - early pressure on d6 in particular which often obliges Black to accept doubled f-pawns - with a degree of solidity missing from, for example, the

main lines of 6.!c4. One fundamental decision White has is where to play his f-pawn. Throughout - Games 3-5 - I have opted for f4 based systems, in spite of a current surge of fashion for f3 followed by a kingside pawn storm. I strongly suspect that these will better stand the test of time, and that the strategic ideas are also easier to explain and to grasp. It also seemed important to create an internally consistent repertoire - it is easier to get a feel for playing these positions if 'mix and match' is kept to a minimum.

Having co-authored a rather substantial work on the Richter-Rauzer in 1 998 , three main questions sprang to mind as I embarked on the project of constructing a repertoire for White that is effective, efficient and internally coherent.

1 ) How much have the fundamentals of the theory changed in 6 years?

2) How, even allowing for the advantages of the repertoire format, will it be possible to be able to condense such a chunk of theory into a neatly proportioned chapter?

3) Lastly, even though my previous work took a 'neutral' perspective, a certain emotional attachment to the Black cause was inevitable, and although I have played both sides, my ailing memory has rendered outings with l .e4 something of a treat. How would I feel advocating the White side, trying to do damage to the Classical Sicilian?

1 1 2 The Classical

Well, the reader will ultimately have to judge how I have risen to the task. The answer to question 1 varies greatly with different lines - the system with 7 . . . a6 and S . . . h6 covered in Game 24 has been entirely transformed by a new and dangerous attacking idea. By contrast, the changes to theory in Games 25-27 are essentially those of detail.

Never forget either, fashion in opening theory moves in mysterious ways. This we shall witness right here in game 23. After 6 . . . �d7 I sense a drift ftom 7.il.e2 towards the more voluminous theory of 7. Wid2. I see no reason - let's keep things simple!

Game 23 Kotronias - Schwartz Philadelphia 2000

1.e4 c5 2.�f3 d6 3.d4 cxd4 4.�xd4 �f6 5.�c3 �c6 6 . .tg5

6 . . . i.d7 The most frequently encountered,

but by no means the only alternative to the main line 6 . . . e6 of Games 24-27. The others I would like to consider in approximately ascending order of importance:

a) 6 . . . g6? ! has a dubious reputation and this is pretty well deserved. In fact after 7.il.xf6 exf6 Black's hopes of dynamic play to compensate his structural damage are not completely without foundation, but with careful handling should be insufficient. For example, S .il.c4 il.g7 9 .lLldb5 0-0 1O .Wixd6 and now:

al ) 1 O . . . f5 1 1 .0-0-0 Wig5t 12.f4 Wih6 13 .'i:t>bl fxe4 14 . lLlxe4 il.g4 1 5 J :!de 1 gadS 1 6. lLlf6t 'i:t>hS 17 . lLlxg4 Wih4 I S .Wic7 Wixg4 1 9 .1Lld6 with a clear plus in Goloshchapov - Chernikov, Moscow 2002.

a2) 1O . . . Wia5 ! ? I 1 . Wic7 Wib4 12 .il.b3 f5 1 3 .0-0-0 fxe4 14. lLl d6 also looks good for White, since the complications arising from 14 . . . il.xc3 I S .bxc3 Wixc3 16 .lLlx£7 'i:t>g7? ( 1 6 . . . Wial t 17. 'i:t>d2 Wid4t I S . 'i:t>e1 Wib4t 19 .'i:t>f1 'i:t>g7 20. lLl g5t Wie7 is better but still unpleasant for the defender) 17. lLl gst 'i:t>h6 I s .Wixh7t 'i:t>xgS 19 .h4t 'i:t>f6 20J �d6t 'i:t>eS 2 1 .Wig7t and wins.

b) 6 . . . WiaS is well met by 7.il.b5! . (In my view this is considerably stronger than the also popular 7.il.xf6 gxf6 s. lLl b3 Wi gS! for one good reason. White's pieces ensure that Black's queen will not become active by crudely but effectively blocking


off the more enticing squares.) 7 . . . !d7 B . ttlb3 �b6 (Or B . . . �dB 9.!xf6 (9.f4!?) 9 .. . gxf6 10 .�h5 E:gB l 1 .g3 E:g5 12 .�e2 a6 13 . .td3 e6 14 .f4 E:g7 1 5 . �h5 �b6 1 6.0-0-0 0-0-0 17 . �bl �bB I B .�h4 when White has a pleasant version of a structure which will become very familiar - Topalov - Corral Blanco, Spain 2000.) 9.!xf6 gxf6 1 O. ttl d5 �dB 1 1 .0-0 (also 1 1 .�h5 e6 1 2. ttl e3 a6 13 .!e2 �c7 14.0-0-0 !e7 1 5 .�b l 0-0-0 1 6.f4 E:df8 17. E:d3 �bB I B . E:hdl !cB 19 .a3 left Black passive in S. Nikolic - Gufeld, Kislovodsk 1 96B. With this characteristic Rauzer doubled f-pawn structure the knight on e3 is rather well placed for restraining any counterplay.) 1 1 . . . !g7 12.c3 0-0 13 .�h5 a6 14 . .txc6 bxc6 1 5 . ttl e3 c5 16 . E:adl E:bB 17. E:d3 h6 I B .E:fdl E:b7 1 9 .ttlxc5! with decisive advantage in Korneev - Lopez Guerrero, Malaga 200 1 . This time the quality of the knight on e3 rather speaks for itself!

c) 6 . . . �b6!? In common with other versions of this early queen sortie, this has enjoyed a good deal of popularity in the last few years. Clearly it raises a number of transpositional issues, especially as I am keen to avoid those main lines of the 7 . . . !e7 Rauzer (see games 26-27) in which White plays an early ttl b3. These could easily be reached here by 7. ttlb3 e6 B. � d2 !e7 9 .0-0-0 0-0 etc. A bit undecided how best to combat this move, I will mention two possibilities, the first ambitious and relatively unexplored, the

second positional, but hopefully retaining some bite, and shifting the battleground to a critical structure which will recur throughout the chapter:

c 1 ) 7 . .te3!? still seems to me to be interesting six years on, but it has not really found very many takers, despite further endorsement from NCO! Critical can only be 7 . . . � xb2 {not 7 . . . ttl g4 B .ttld5 ttlxe3 9.fxe3 �a5t 1 O.b4! winning material, while 7 . . . � c7 B. ttl d5 looks nice for White. 7 . . . a6 cannot be a major test either. White had a pleasant position after B .�d2 ttlxd4 9 . .txd4 �c6 10 .!d3 e5 1 1 .!e3 !e6 1 2.f4 exf4 1 3 .!d4 !e7 14 .�xf4 0-0 1 5 .0-0-0 E:acB 16 .E:hfl ttl d7 17. ttl d5 !dB I B .c3 in Ivanchuk - Miroshnichenko, Antalya (EuropeanCh) 2004) B . ttl db5 �b4 9.!d2 �c5! {9 . . . ttl e4? 1 O .a3 is a catastrophe for Black, which recently befell so highpowered a victim as the young Georgian star Baadur Jobava. At least he had the

1 14 The Classical

good grace to resign forthwith!) 1O .i.e2! ( I O.i.e3 repeats, while 1 0. 'tJ c7t @d8 1 1 . 'tJ xa8 'tJ g4 gives Black excellent compensation since 12.'1& e2 loses to the acutely embarrassing 12 . . . 'tJd4 13 .1&d3 'tJ e5 !) 1 0 . . . 1&b6 l U ' ib l 1&d8! (an improvement over 1 1 . . . 'tJe5? 14 . .te3 1&a5 1 5 .l''i b3 g6 16 . E1a3 1&d8 17.i.xa7 'tJ ed7 1 8 .f4! Balashov - Petrienko, Voronezh, 1 987) and now I think the simple 1 2 . 'tJ d5 'tJxd5 1 3 .exd5 'tJ e5 14.0-0 offers good compensation since Black has no straightforward means of developing. This still awaits a practical test, although to be fair it has been Black as much as White who has steered clear of it.

c2) 7. 'tJ b3 e6 8 .i.xf6!? Before Black can play . . . .te7. I have to confess, I am not convinced that these positions should give White a theoretical plus and hence here and later in the book I shall always try to offer an alternative. However, it is invaluable to learn how to handle the structure with the doubled fpawns and, in addition, White's position does seem easier to handle in practice. 8 . . . gxf6 9 .ygd2 a6 10.0-0-0 i.d7 l 1 .f4 0-0-0 12 .@bl @b8 1 3 . .te2 h5!? There is plenty of scope here for move order flexibility. However, I suspect that it is a good sound instinct to answer .te2 thus. One example of omitting this precaution - 1 3 . . . i.e7 14 .i.h5 i.e8 1 5 . E1hel E1c8 1 6 .'tJd5 ! exd5 17.exd5 'tJ e5 1 8 .fxe5 fxe5 19 . E1fl f6 20.i.xe8 E1hxe8 2 1 .ygd3 with

a great superiority on the light squares. Nataf - Thorhallsson, Reykjavik (rapid) 2003. See also the superb game Almasi - Damljanovic, given under the note (a to Black's 7th in game 26.

14 . E1hfl i.e7 1 5 .E1f3 E1dg8 16 . .tfl E1g4 17.a3 i.c8 1 8 . 'tJ a4! 1&c7 1 9 .YGfl b5 20. 'tJc3! Instructive, both how White forces this weakness, and the ferocity of the attack, which he builds thereafter. 20 . . . 1&a7 2 1 .ygel E1hg8 22.a4 b4 23. 'tJ a2 a5 24.i.b5 i.b7 25 .i.xc6 i.xc6 26. 'tJ xa5 ! A nice combination which strips Black's king bare. 26 . . . ygxa5 27. 'tJxb4 @b7 28. E1b3 YGxa4 29. 'tJxc6t @xc6 30.1&c3t @d7 3 1 . E1b7t @e8 32.ygc8t i.d8 33.b3 1-0 Alekseev - Kiselev, Suetin (mem) Tula 2002.

d) 6 . . . a6 is by no means bad, but since Black is often liable to meet 7 .1& d2 with 7 . . . e6 it has no independent significance in the majority of cases. However Black does have a distinctive idea, albeit a rather inferior one, in 7 . . . 'tJxd4. This makes a certain sense when White has already played 7.1&d2. Indeed just such logic underlies the preference for 7.i.e2!? in the main game. However, after 8 .ygxd4 the follow-up 8 . . . e5 cannot be recommended. As usual the price to pay for this weakening of d5 is especially high in the Rauzer, and the inconvenience to White's queen is scarcely significant in lines like 9 .1&a4t! i.d7 1 0.i.xf6!? gxf6 ( I O . . . i.xf6?! allows the shot 1 1 .i.b5! 1&d8 1 2 . .txd7t ygxd7 13 .YGb3 .te7

14 . ltld5 i.d8 1 5 .0-0 0-0 16 .E:fdl and White has one of those examples of minor piece superiority which are normally only seen in textbooks. Kotronias - Kovalev, Debrecen, 1 992.) 1 1 . \Wb3 b5 12 .i.e2. White will follow up with ltl d5 with a safe plus. 7.�e2!?

Exceptionally, since Black cannot switch to . . . e6 without complication (see 'c' below) , there is mileage here to the idea of castling short with i.xf6, ltl fS and ltld5 in the air. This plan is no longer high fashion, but still seems to me to have a lot of bite.

7 . . . a6 Four alternatives, one rather

questionable, two of great importance, plus one that appears to be a slightly inferior way of trying to reach positions that could be arrived at a good deal less painlessly!

a) 7 . . . \Wb6 is rather poorly timed

as White has 8 . ltldb5! E:c8 9.0-0 a6 1 0.i.xf6 gxf6 l 1 . ltld5 \Wd8 12. ltl bc3 e6 1 3 . ltl e3. Again one of the virtues of the move ltl d5 is that even if the knight is forced to retreat, this is rather a good square. 1 3 . . . b5 14.a4 b4 1 5 . ltl a2 \Wb6 1 6.c3 a5 17 . ltlc4 \Wc5 1 8 .ltlc 1 ! with strong pressure against d6. Vogt -Mascarinas, Polanica Zdroj 1 977.

b) 7 . . . E:c8 is probably just an attempt to reach the critical positions considered under 'd' below. It is justworth mentioning because after 8 .0-0 ltlxd4 9 .\Wxd4 \Wa5 . I rather suspect that as well as 1 0 .i.xf6 gxf6 I I .a4 returning to line 'd' , White can also consider 1 0.i.e3!? i.c6 and then some solid move like 1 1 . E:adl , when the black pieces somehow look a bit wayward.

c) 7 . . . e6!? is rather a plausible response to a developing move which in general terms may be viewed as rather modest. It has moreover been greatly strengthened in recent times by the discovery that there may be a nasty surprise in store for White if he simply tries to head for the classic 'Rauzer structure' with the doubled f-pawns. After 8.i.xf6?! , a young Dutchman overturned the previously unquestioned verdict of theory by playing 8 . . . \Wxf6! and after 9. ltldb5, far from the promised '± ' White was confronted with 9 . . . O-O-O! in D. Mastrovasilis - Berkvens, European (U20) Patras 200 1 , and shied away from 10 .ltlxd6t <Jtb8 l 1 . ltlc4 i.c5 12.0-0 ltl e5! , which indeed would promise Black excellent compensation on

1 1 6 The Classical

the dark squares. Of course after others 9. ttJ bS makes little sense and Black was quickly better. It is often the hallmark of a good novelty that it seems startlingly obvious in retrospect!

Therefore White needs to look elsewhere. Quiet moves like 8.0-0 or even 8. 'Wf d2 are unobjectionable, but 8 . ttJdbS!? seems more promising. E.g. 8 . . . 'Wfb8 9.a4! a6 1 O .ttJa3 'Wfc7 l 1 .'Wfd2! (less common, but more incisive than the routine 1 1 .0-0. White should force the issue by piling immediate pressure on d6) 1 1 . . . .te7 12J=1dl ! and now:

c 1 ) 1 2 . . J=1d8 is natural but a bit passive. White can exploit the weakness of b6 with gain of time by 13 . ttJ c4 .tc8 14 . .te3! 'Wfb8 l S . ttJ b6!? ttJd7 1 6.0-0 0-0 17.f4 ttJ cS 1 8 . .tf3 with a modest but pleasant initiative in Spassky - Hort, Moscow 1 97 1 .

c2) 1 2 . . . ttJ eS ! ? played in Lautier -Milov Cap d' Agde 2002 is a more active

defence inviting serious complications. After 1 3 .f4 ttJg6, instead of the slightly lackadaisical 14.0-0 White should have played 14. ttJ c4! dS!? (All but forced. Not 14 . . . O-O? l S .eS dxeS 16 . .txf6+-, while 14 . . . .tc6 l S . .txf6 gxf6 1 6.f5! is also very unpleasant.) I S .exdS ttJ xdS 1 6. ttJ xdS exdS 17. ttJ e3!? (Wells - better I think than 17 .'WfxdS .tc6 1 8 .ttJd6t .txd6 19 .'Wfxd6 'Wfxd6 20Jhd6 f6! 2 1 .f5 ttJ e7! 22 . .tf4 ttJxf5 when Black is quite OK.) 17 . . . .txgS 1 8 .fxgS .te6 1 9 .0-0 with an unusual position and structure, but I think slightly better chances.

d) 7 . . . 'Wf as!? is arguably the single greatest challenge to White's set-up.

With 7 . .te2 already on the board, the antidote which was so effective against the premature 6 . . . 'WfaS makes little sense here, and this helps to ensure that Black's queen will enjoy some role along the 4th rank - perhaps with the move . . . 'WfgS , or else utilising the fact that a well-timed

Experts VS. the Sicilian 1 1 7

exchange on d4 can be followed up with the tempo gaining . . . �c5. Best in my view is B .�xf6 gxf6 9.0-0! (9. ltlb3 'Wg5 is quite playable for Black) and now Black has tried several moves:

dl ) 9 . . . 0-0-0? cuts off the queens retreat and causes her severe discomfort after 1 0 .ltld5 e6 l 1 . ltl b3! 'Wa4 12. ltl c3 'Wb4 13 .�b5 d5 14.a3 'Wd6 1 5 .exd5 ltl e5 16 .�xd7t �xd7 17.dxe6 �xe6 I B .'Wh5 with obvious advantage in G. Shahade - Thorhallsson, Elbow Beach Club 200 1 .

d2) 9 . . . :BgB 1 0. ltl d5 ! (a key move, and also the main response to 'neutral' Black options on move 9) 1 O . . . �h3 1 1 .�f3 ltl e5 1 2. ltl e3! ( l 2. ltl f4 �d7 achieves little, and is usually only used for purposes of repetition) 12 . . . �h6! (of course 12 . . . �d7 is rather pliant, and White has a dear plan in 13 .�h5! e6 14.f4 ltlc6 1 5 .f5 with a dangerous initiative in Romero Holmes - Cifuentes Parada, El Vendrell I 996) 1 3 .ltldf5 �xe3 14. lLlxe3 'Wd2! (safer than 14 . . . 0-0-0 1 5 . cj;>hl �d7 1 6. lLl d5 :BdeB 17.a4 f5 I B .b4 'WdB 1 9 .exf5 �xf5 20.a5 Borriss - Sherzer, Santiago (U-20) 1 990, when it is White's queenside play which looks much the more serious.) 1 5 . cj;>hl 'Wxdl 1 6.�xdl �d7 1 7.f4 lLlc6!? (if 17 . . . lLlg4 I B .ixg4! ixg4 1 9 .f5, Black can extricate his bishop by means of 1 9 . . . ie2 20. :Bf2 ib5, but the white knight on d5 will still be a fine piece securing some advantage) I B .lLld5 :BcB 19 .c3 and although the

players agreed a draw here in Ebeling -Krogius, Jyvaskyla 1 99 1 , Black is rather short on counterplay.

d3) 9 . . . lLlxd4 10 .'Wxd4 :BcB (But not 1 0 . . . 'Wc5 ? 1 1 . 'Wxc5 dxc5 1 2. ltl b5! Short - Anand, Amsterdam 1 992)

when White has to choose between two interesting options:

e3 1 ) l 1 . lLl d5. An interesting pawn sacrifice courtesy of Beating the Sicilian 2, which initially breathed life back into the then ailing 9.0-0. In general White was able to generate substantial play for a pawn, but the defender eventually found an antidote in the form of 1 1 . . . 'W c5 12 .'Wd2 ( l 2.'Wxc5 ?! :Bxc5 leads to an ending in which the bishop pair is likely to really come into its own) 12 . . . �xc2 13 .�e3 ih6! ( l 3 . . . 'Wc5 14 .�b3! ig7 1 5 .�xb7 e6 1 6.�b5! 'Wxb5 17.�xcBt! is an elegant trick which Vassilios Kotronias has had the opportunity to reveal twice!) 14 .'Wxh6 �xe2 1 5 .'Wg7 :BfB 16. :Bacl

1 1 8 The Classical

:1!xc1 17 . :1!xc1 i.c6 1 8 .tiJc7t 'it>d7 19 .�xfS �xe4 20.�g7 'it>xc7 and Black's queen will return to g6, and White will be hard pressed to avert an ending in which his opponent enjoys ample compensation for the exchange.

e32) l 1 .a4! This now appears the more promising. If Black does not prevent it then there is a very simple but rather effective positional idea of exchanging light-squared bishops on b5. If he does then the pawn sacrifice is enhanced. 1 1 . . . � c5 12 . � d2 and again a choice:

e32 1 ) 1 2 . . . i.g7 13 .i.b5 f5 14. tiJ d5! i.xb5 1 5 .axb5 fxe4 1 6.�g5 i.f6 17 .tiJxf6t exf6 1 8 .�xf6 �e5 19 .�xe5t dxe5 20. :1!xa7± Hracek - Heberla, Czech 2004.

e322) 12 . . . h5 1 3 .i.b5 i.h6 14 .�d3 e6 1 5 .e5! clearing the e4 square was no less unpleasant for Black in Gallagher -Weindl, Switzerland 2004.

e323) 12 . . . a6! ? At this point I wrote in 1 998 "the interpolation of a4 and . . . a6 benefits White here in view of the weakened b6-square." Good authorial waffie, but I had no idea of quite what specific importance this would turn out to be!

In fact, in the case analogous with the 'refutation' of 1 1 . tiJ d5 given above (see e3 1 ) the difference is probably decisive, although so far as I can see this has not yet been mentioned in the literature. After 1 3 . tiJ d5! �xc2 14 .�e3 i.h6 1 5 .�xh6 �xe2

16 . � g7?! is no more promising than it was above, but there is an additional possibility: 1 6. tiJ b6! :1!c7 17.�g7! :1!fS 1 8 .tiJxd7! (At this moment when Black must recapture with the rook. The king should be kept in the centre, and weakening the back rank wins time too.) 1 8 . . . :1!xd7 19 . :1!ac1 :1!d8 20.�xh7! (preventing . . . � xe4 which would grant the black queen a route back to g6 and effective defence) and White threatens to double on the c-file, while Black will have great difficulty playing with any pieces other than the queen. If this holds up, it should make a major contribution to bolstering the reputation of 7.i.e2. 9.c!LJf5!?

Exploiting the absence of the move . . . e6, this at the same time prepares an interesting pawn sacrifice, as we shall see. 9 • • • 'iNa5

Extraordinarily, Black's main alternative here is the 'un-developing move' 9 . . . i.c8! ? Somehow, even if the move makes sense,


White can take some encouragement from the fact that his opponent is that desperate to remove the knight from f5. After 10.a4 e6 1 1 . lDe3 b6 12 .0-0 i.b7 White has a choice of decent plans. 13 . VN d2, followed by Eiad1 and lD c4, or perhaps even more simply as in Stefansson - Zubarev, Las Vegas (Wch) 1 999, just 13 .@hl ! ? i.e7 14.f4. Now Black did himself no favours by 14 . . . O-O? since the attack almost plays itself and after 1 5 .i.c4 @h8 16 .VNh5 VNe8 17.f5 lD e5 1 8 .i.b3 granting the e4 square with 18 . . . d5? was already the last straw. 1 9.exd5 i.c5 20. Eiae 1 Eig8 2 1 . lD e4 i.xe3 22. lD xf6 1-0. It seems unlikely that the time consuming 9 . . . i.c8 will solve Black's problems. After all, as I have said before, the knight is a good piece on e3! 10.0-0 hf5 1 l.exf5 Wxf5 IV�d5

12 . . . �d8 Of course Black has to deal with lDc7t

and the alternative 12 . . . Eib8 also scarcely leaves castling realistically on the agenda.

One very powerful attacker then built White's initiative effectively with 13 .i.g4 VNg6 14 .i.h5 'Wig7 1 5 .Eie 1 Eig8 1 6.g3 f5 17.c3 @d7 1 8 .b4! e6 1 9 .1Df4 i.e7 20.a4 i.f6 2 1 .b5 in Vitolinsh - Grokhotov, USSR (ch) 1 975. 13.J.g4 Wg6 14.i.h5 Wg7 15J�el e6 16.�f4 �d7 17.c4!

There is no set formula for handling the White side here. His strengths are his opponent's lack of effective development and his own active minor pieces. Given the positional costs of the move . . . e5, the knight also enjoys a de facto outpost on f4. Still, as in the example in the note above, it is a queenside pawn storm that gives this initiative its extra bite. 17 . . . l:i:d8 18.b4 Wg5?!

It is not easy to give Black good advice here, although he might have preferred the immediate 1 8 . . . lD e5. What is certain is that this 'changing of the f7 guard' is not viable, as Kotronias' excellent combination shows. 19.93 �e5 20.i.xf7! �xf7 21.�xe6 Wf5 22.Wa4t �c8 23.h5!

The hallmark of a strong attacking player. White is in no hurry to recoup his material, but rather uses the powerful position of his knight on e6 to spearhead a direct onslaught. 23 . . . �e5

Or 23 . . . lDg5 24. lDxg5 fxg5 25.bxa6 @c7 26. Eiab 1 b6 27.VNb3 'Wia5 28.'Wif3! and White invades decisively. 24.bxa6! Wxe6 25.axh7t �c7

1 20 The Classical

25 . . . �xb7 is met simply with 26J'1eb l t �c8 27.VfJ.a7 mating. 26.YlYa5t i>d7 27.YlYxdSt i>xdS 2S.bSYlYt YlYcs 29.:aabl �d7 30.YlYa7?!

30. :ae8tL 30 • • • b5 31J:�b7 gb7 32.gebl h4 and Black decided he had had enough, as 33. VfJ.a5t �e8 34. VfJ.f5 is crushing. 1-0

Vintage Kotronias, in his element in such attacking positions. White's compensation in any case looks very believable, and my hunch is there will not be too many takers for the cause of 7 . . . a6.

Game 24 Balasbov - Makarov Smolensk 2000

1 .e4 c5 2.�a �c6 3.d4 cxd4 4.�xd4 �f6 5.�c3 d6 6.i.g5 e6 7.YlYd2 a6 s.O-O-O b6

9.�xc6!? From the author's standpoint, this has

been a very tough decision. At the time when The Complete Richter-Rauzer came out ( 1 998) I would have had no hesitation at all in building a repertoire around the flexible and far from innocuous retreat 9 .ie3. Indeed, at that stage, the line to be recommended here was little more than a mildly eccentric backwater. Now it is the height offashion! I have to admit that there is still a part of me that rebels against this exchange on c6. Indeed there should be! It strengthens Black's centre, enables him to effect the move . . . d5 under unusually secure circumstances, and opens the b-file for action that potentially could leave the white king most uncomfortable.

So what can White boast in return? Well, let us note first that the b-file problems are for the moment likely to be largely academic. Not only is the b8 square well covered, this is unlikely to change for a long time since the f4 bishop is pretty difficult to dislodge. In fact this is a great piece - White enjoys a qualitative lead in development as well as a quantitative one. The question is: Can Black neutralise this initiative? In essence, it boils down to dynamic vs. static features, and this usually makes for an entertaining scrap. 9 • • • bxc6 10.1f4 d5 1 l .YlYe3!

Again White is able to find aggressive squares for his pieces, which at the same time restrict his opponent's mobilisation - an unusual and very effective dark


square influence i s being exerted along two sweeping diagonals. Still, advantages in development have a habit of being rather transient, and Black to move now has a very fundamental choice of ways to get his bits into the game. 1 l . . . YlYa5!?

Interestingly, even as my opinions about the variation as a whole have fluctuated wildly, my belief that this is the most promising try has stayed quite stable. Let us first take a look at the others:

a) 1 1 . . . i.e7 represents Black's simplest mode of development, but the potency of the attack which White can generate is in turn the best advertisement for his dynamic chances. As usual here White plays 12 . i.e2, directed against . . . tLl h5, when Black can try:

a l ) 1 2 . . . 0-0 13 .h4! ? (the older 13 .�g3 Wh8 14 .i.c7 Y;l.Id7 1 5 .i.e5 Y;l.Ib7! 1 6.exd5 cxd5 17J'1d4 tLl e8 1 8 .i.d3 i.f6 of Nisipeanu - Ivanchuk, Las Vegas (Wch)

1 999 is to my mind less convincing) 1 3 . . J��e8 14 .�g3 Wh8 1 5 .i.c7! Y;l.Id7 1 6.i.e5 i.fB. Fressinet - Bacrot, France (ch) 2000 and now 17.�f4!? tLlg8 1 8 .i.h5! Eia7 1 9 .94! - Gofshtein, would have presented Black with huge problems.

a2) 1 2 . . . tLld7 has aspirations to control e5 and maybe have the liquidating . . . i.g5 available too. White should try 1 3 .h4! when after 1 3 . . . �b6 14 .�g3 d4?! 1 5 . tLl b l �a5 1 6. Eixd4 e5 17.i.d2 �b6 1 8 . 1".k4 he already had a significant advantage in Khalifman - Xu Jun, China-Russia Shanghai 200 1 . Black can try a couple of improvements, but 14 . . . g6 which has been suggested looks very dangerous too in view of 1 5 .exd5 cxd5 16 . Eixd5! exd5 17 . tLld5 with every prospect of winning back material with a strong initiative. Also 1 3 . . . 0-0 failed in dramatic fashion to 14.exd5 cxd5 1 5 .i.xh6! in Bauer - Chabanon, France (ch) 2003 since if 1 5 . . . gxh6 1 6.�xh6 tLl e5 ( l 6 . . . Eie8 17. Eid3 i.d6 1 8 .f4!) then 17. tLl e4! dxe4 (otherwise 1 8 .tLlg5 will be too strong) 1 8 . Eixd8 Eixd8 19 . Eih3! when White has regained material and retains a powerful attack.

b) 1 1 . . . i.b4 does not overly impress me. Since 1 2.a3 looks good, it is not even a legitimate move order for transposing to the main game. White has 12.a3!? i.a5 ( l 2 . . . i.xc3 13 .Y;l.Ixc3 tLlxe4 14.Y;l.Ixg7 Y;l.If6 1 5 . Y;l.Ixf6 tLlxf6 1 6.i.e5! We7 17.g4 favours White) 13 .i.e2 0-0 14.e5 tLld7

1 22 The Classical

( l 4 . . . i.xc3 1 5 .�xc3 'Ll e4 16 .�e3 f6 17.£3 fxe5 18 .i.xe5 �g5 was better for White after the queen exchange in Degraeve - Apicella, France 1 999, but 1 9 .f4! �e7 20.�£3 could have heralded a still more punishing response.) 1 5 .�g3 �c7 1 6. ghe 1 f6 17.i.xh6 �e7 1 8 .i.h5 i.xe5 19 .i.f4 i.xf4t 20.�xf4 e5 2 1 . �h4 gd8 and Black's strong centre gave him counter-chances in Mainka - Kritz, Germany (ch) Hoeckendorf 2004. However, while . . . f6 can be quite a potent idea, Black really seems to me to lack a 'plan b' . So why not play a move which really acts as a deterrent to this pawn break. After 1 6.i.d3!

1 6 . . . f6 17 . � g6 is really too dangerous, while 'quiet' moves like 1 6 . . . �e7 17. ghe1 I!;>h8 allow 18 . �h3! and if now 18 . . . f6, the sacrifice 1 9 .�xh6 gxh6 20.�xh6t I!;>g8 2 1 . ge3 i.xe5 22.f4! is very strong.

c) 1 1 . . . � e7 also makes a rather artificial impression, especially when development

deficit is Black's main challenge. Chandler - Bellin, Birmingham 2000 was typical. 1 2.exd5 cxd5 13 .i.e2 �a7 14.�g3 d4 1 5 .'�M3 i.d7 16 .�xd4 i.c5 17.�e5 i.xf2 1 8 .i.£3 gc8 1 9. 9d3 i.b6?! and now White could have caused quite severe problems already with 20. 'Ll e4! . 12.J.e2! i.b4?!

It is easy to be wise after the event. In putting the question to this move, it is worth reminding ourselves that it is only the extraordinary energy of White's solution that reveals its deficiencies. The problem is that after the coming exchange on c3, White's dark squared bishop (his pride and joy in this variation) will attain new heights and Black's inability to contest it will compensate for a pawn and some inconvenience to the white king. Two alternatives, the second of which is of great importance and for a time gave me serious pause for thought before advocating White's hyper-aggressive setup:

a) 12 . . . � c5 is all about driving away the queen in order to play . . . d4. However, positional problems persist after 1 3 . � g3 d4 14. 'Lla4 �b4 ( 14 . . . 'Llxe4? 1 5 .�f3 �d5 1 6.c4!) 1 5 .�b3! �xb3 1 6.axb3 'Lld7 17. gxd4 e5 1 8 . gxd7! �xd7 19 .i.xe5 i.e6 20. gd1 i.e7?! (20 . . . gd8 2 1 .�xa6! looks grim too) 2 1 . i.d6! a5 22.e5!+- h5 23.�f3 when White had a terrific grip, Balashov - Kiselev, Russia 2000.

b) 12 . . . dxe4!? is a tough nut to crack.


b l ) I have spent a long time trying to make 13 .ic4 work ( 1 3 .lLlxe4 lLldS 14JhdS cxdS I S . lLl d6t is also not without some venom, but is at the same time unnecessarily speculative) but had to bow to the inevitable after the excellent reply 1 3 . . . VNf5! 14.B ( 14.g4 lLlxg4 I S .VNb6 11*fxf4t 1 6.';!;>bl id7 17Jhd7 @xd7, Nataf, - Gershon, Bermuda, 1 999 is fun, and after 1 8 .VNb7t VNc7 1 9 J%dIt id6 20.ixd6t @xd6 2 1 . lLl e4t etc. White should have enough to draw, but not more) 14 . . . icS! I S .VNd2 0-0 1 6.fxe4 lLlxe4 17 . lLlxe4 VNxe4 1 8 .id3 11*fa4 and Black holds the balance. Again 1 8 .ixh6!? should lead to no more than a draw by perpetual.

b2) 1 3 .11*fg3! is much more to the point, threatening 14 .ic7 and monitoring the g7 pawn too. The real conceptual breakthrough is that White is even willing to 'mend' his opponent's compromised structure in the interests of maintaining

the momentum of his initiative. After 1 3 . . . lLldS! 14 . lLlxdS ( 14 .ieS lLlxc3 I s .ixd3 11*fgst) 14 . . . cxdS I S .@b l id7!?

he can try: b2 1 ) 1 6.B?! which was recommended

by Tsesarsky, who offered only the rather compliant 1 6 . . . exB 17.ixf3, when indeed it is plausible that the opening of another file piles more woe upon the defender. Unfortunately, the liquidation which accompanies the very natural 1 6 . . . ibS! 17 J :!he l ixe2 1 8 Jhe2 exf3 19 .11*fxB ie7 looks just fine for Black. White adapted sensibly in ZaragatskiRau, Germany UI8 (ch) Willingen 2003 with 20.c4 0-0 2 1 .cxdS �fd8 22. �ed2 exdS 23. �xdS VNb6 24. �xd8t and a draw was agreed.

b22) 16 .ieS! is strongly preferable, and consistent with White's fundamental strategy: keep the black pieces at home! Lines like 16 . . . �g8 17.f4!? ibS 1 8 .ihS!

124 The Classical

g6 1 9 .f5!? look very promising, but the whole thing needs practical tests. 13.i.e5! ,ixc3

1 3 . . . dxe4 is less ambitious, arguably less consistent and also probably falls short of equality although the young Chinese talent Bu Xiangzhi's patronage gives pause for thought. White retained a structural plus after 14 .�xf6 gxf6 1 5 .�xe4 <JJ e7 1 6J �d3 �a7 17.a3 �xc3 l S . �xc3 in Degraeve - Lerner, Koszalin 1 999, while 14 .�g3!? �xc3 1 5 .�c7 �xb2t 16 . <JJxb2 �b4t 17 . <JJal ttJd5 I S . �xg7 �f8 19 .�d4 �d7 20.�g3 �gS 2 1 .c4 c5 22.�b2 ttJ b4 23.a3 ttJd3 24.�xd3 exd3 25 . �xd3 was a more punchy route to a good position in Ponomariov - Bu Xiangzhi, Lausanne 200 1 . 14.,ixc3 �xa2

15.,td3!? The priority of course is that Black

should not be able to land a knight on e4, when his counter-chances are real

indeed! The huge power of the bishop on c3 - especially in the context of opposite coloured bishops - in conjunction with the weakness of g7 are the grounds for White's tremendous idea. I have to admit, I have found no flaw with Balashov's play, but in later encounters White has demonstrated alternative methods of causing trouble too:

a) 1 5 .�f3 0-0 1 6.b3! dxe4 1 7.�xe4 ttJd5 l S .�xd5 exd5 ( 1 S . . . cxd5 19 . <JJd2! �a3 20.�d4) 19 . �he 1 �f5?! ( 1 9 . . . f6! is not so clear.) 20.�b2 �g6? 2 1 . �c3 f6 22. �d4 1-0 Chanoine - Santo Roman, Paris 200 1 is obviously striking. While Black could have defended much better, the potency of opposite coloured bishops for the attacker was very much in evidence.

b) 1 5 .f3!? 0-0 1 6.g4 �dS ( 1 6 . . . dxe4 17.fxe4 e5 l S .b3! and capturing on g4 is too risky, although White will enjoy a serious kingside attack in any case) 17.g5 d4?! ( 17 . . . �a1 t l S . <JJd2 �dSt 19 .�d3 �a4 20.gxf6 exd3 2 1 .cxd3 �h4! 22.fxg7 also clearly holds dangers for Black, but no clear win is in sight) l S . �xd4 �xd4 19 .�xd4 �al t 20. <JJd2 �xh1 2 1 . �dSt <JJh7 22.gxf6 �b7 23.�e7 c5 24.fxg7 � g2 (24 . . . <JJgS is tougher, but White is still winning after 25 . �h4 <JJh7 26.gSt� �xgS 27.�f6 - Boto) 25 .�f8 �g5t 26. <JJ e 1 e5 27.�xe5 �xf8 2S.gxf8� �xe5 29.�xf7t �g7 30.�xg7t <JJxg7 3 1 . <JJf2 1-0. E. Hossain - Mohammad, Bangladesh (ch) 2003.


15 .. dxe4 16.Yllig3!! The star move, and a graphic illustration

of the weakness of Black's dark squares which as we shall see, persists deep into the endgame phase. 16 . . . exd3 17.Yllixg7

17 . . J�g8 17 . . . d2t does not help. Balashov

himself gives the nice line 1 8 . mxd2 It:l e4 t 19 . me3 :B:fB 20. :B:d8t mxd8 2 1 .�xf8t mc7 22.i.e5t mb7 23.�b4t ma7 24.i.c7 and wins. 18.Yllixf6 d2t

1 8 . . . �a1 t 1 9 . md2 �a4 20.me3! - Balashov - also seems to be good. Black can try 20 . . . i.b7 2 1 . :B:xd3 c5, but 22.B! consolidates, and Black's dark square misery continues unabated. 19.ebxd2 Yllid5t 20.ebel YlligSt 21 .Yllixg5 hxg5

2 1 . . J 'hg5 ! ? 22.i.f6 :B:d5 is tougher, although 23.c4 evicts the rook, with some positional advantage guaranteed.

22 . .if6 .ib7 23.M! gxh4 24Jixh4 e5 25.g4 .tf3 26J:�d3 J.e2 27J:�e3 J.f1 28.g5

Detailed coverage of the remainder of this game is clearly beyond our scope given limited space. Suffice to say that the dark-squared bishop remains the star of the show, and Black soon decides that a rook is a fair price to pay to get rid of it! 28 . . . ebd7 29.e4 ebc6 30.ebd2 gad8t!? 31 .ixd8 gxd8t 32.ebc2 gg8 33.gg4 ebd6 34.b3 gh8 35.g6 fxg6 36.gxg6 .ih3 37.gg7 .if5t 38.ebb2 ebc6 39.ga7 ebb6 40.gd7 ebc6 41 .gd2 ghl 42.gg3 a5 43.gg8 a4 44.ge8t ebb7 45.gxe5 gbl t 46.eba3 axb3 47.gb5t ebc6 48.gxb3 gel 49. ebb4 ghI 50.ga3 gbl t 5 1 .ebc3 gel 52.ga6t ebb7 53.ga5 ebb6 54J�b5t ebe6 55.ga2 gel t 56.ebd4 gdlt 57.ebe5 ge1t 58.ebf6 .id3 59.ga6t ebe7 60.ge5t ebb7 6I Jla3 J.e2 62.ga2 1-0

1 26 The Classical

Game 25 Lastin - Spraggett Moscow 2004

1 .e4 c5 v��a d6 3.d4 cx:d4 4.�xd4 �f6 5.�c3 �c6 6.J.g5 e6 7.'flYd2 a6 8.0-0-0 i.d7

A flexible and popular developing move which as we shall see does not yet commit Black to any single set-up. Along with the 8 . . . h6 of Game 24, this is by far the most significant choice. Others tend to be transpositional. The most significant, 8 . . . i.e7, is likely to lead to the main lines of game 27. 9.f4 b5

One of the main lines of the Rauzer over many years, this attempt to create rapid play on the queenside without committing the e7 bishop has enjoyed a fluctuating reputation but has also enjoyed the truly fierce partisanship of a committed group of devotees. Chief

among these are the talented Croatian grandmaster Zdenko Kozul and the wily Yugoslav Branko Damljanovic who have repeatedly shown the enormous power latent in the black bishop pair following the exchange on f6.

Two other moves also deserve detailed consideration:

a) 9 . . . !e7 10. tLl f3!? b5

1 1 .!xf6 and now: a l ) 1 1 . . . !xf6 is not quite respectable.

After 12 . 'flYxd6 i.e7 ( 12 . . J %a7!? is interesting. 13 .e5 !e7 14 .�d2 �a5 1 5 . Wb l !b4 1 6. tLl g5 !xc3 17 .�xc3 �xc3 1 8 .bxc3 tLla5 restricts White to a modest edge, but 14 . � d3! ? is worth a look too) 13 .VNd2 b4 14 . tLla4 tLlb8! ? (after 14 . . J %a7 1 5 .VNe3 �a5 1 6.b3 E:b7 17 . tLl d2 Black's compensation is rather effortlessly neutralised) 1 5 .VNd4 VNc7 1 6. tLl b6 E:a7 17 . Wb l 0-0 1 8 .tLlxd7 tLlxd7 19 .e5 E:c8 20.!d3 tLl c5 was Illescas Cordoba - San Segundo, Madrid


1 997 when 2 1 .f5 ! would have been strong.

a2} 1 1 . . .gxf6 is the main line and introduces yet another version of the doubled f-pawn structure. It bears obvious comparison with the main game. White has withdrawn his knight to 8, while Black's bishop is committed to the relatively modest e7. Indeed there is a sense that the Black position is a little passive in comparison with the 9 . . . b5 lines. After 12 .i>b1 1Wb6 1 3 .f5 0-0-0 14.fxe6 fxe6 1 5 .g3! i>b8 16 .�h3 �c8 17 .1We1 ghe8 1 8 . tLl e2 we reach a parting of the ways:

a2 1 } 1 8 . . . tLle5 1 9. 9n tLlc4 20. tLlf4 �fB (20 . . . tLl e3 2 1 .�xe6 tLlxd1 22. YNxdl ± completely and unacceptably cedes the light squares) 2 1 . YNf2!. Instructive. The exchange of queens will still leave White with enough targets in the Black position. The priority in this strategy should be given to restraint. 2 1 . . . YNxf2 (2 1 . . .YNb7

offered more chances of counterplay, but 22. tLl d2! goes a good way towards nullifying this too) 22. gxf2 f5 23.exf5 tLl e3 (23 . . . exf5 24 tLld4!) 24. ge1 tLlxf5 25. tLlg5 e5 26. tLld5 tLld4 27.�xc8 gxc8 28 .c3 tLl e6 29. tLlxe6 gxe6 30. tLlb6 1-0. Bruzon - Molander, Santo Domingo 2003. A snappy victory for the talented young Cuban.

a22} 1 8 . . . j,fB 1 9 .9n �g7 20. tLl f4 and now:

a22 1 ) 20 . . . j,h8 2 1 .c3 ge7 was Adams - Timman, Wijk aan Zee 2004. White played 22. gf2(?!} which granted time for 22 . . Jl:de8 so that 23. tLld4 could be met with 23 . . . tLla5! and some counterplay. As Adams himself points out the immediate 22. tLld4! was better since 22 . . . tLlxd4?! 23.cxd4 is very pleasant for White, while 22 . . . gde8 23. tLlxc6t 1Wxc6 24. gf2 again represents a success for White's strategy of constraint.

a222} 20 . . . YNc5 2 1 .c3 f5 (2 1 . . . j,h6 22. tLld4 �xf4 23. gxf4! makes limited sense as 23 . . . e5 can be met with 24. tLl b3! and 25. gh4 with substantial positional plusses) 22.exf5 exf5 23. YNf2 b4 24.cxb4 tLlxb4 25.a3 tLl c6 26. gc1 YNxf2 27. gxf2 tLld4 28 . gd1 tLlxf3 29. gxf3 and White retained a small advantage based on his opponent's substantially damaged structure in Kolev - Damljanovic, Skopje 2002

b} 9 . . . h6!? is also interesting. Black prepares . . . g5 and control of the e5 square. Interestingly, recent attention

1 28 The Classical

has centred on White refusing to return the pawn lightly. The resulting positions are, as we shall see, full of tension. 10 .i.h4 g5 ! ? (Rather than 10 . . . lLlxe4?! 1 1 . 'lWe I ! lLlf6 ( 1 1 . . .g5? 12 . lLlxe4 gxh4 13 .�c3 is a major accident) 1 2. lLl f5 �a5 1 3 .lLlxd6t i.xd6 14Jhd6 �c7 (or 14 . . . 0-0-0 1 5J'!dl ! 'lWc7 1 6.'lWf2 lLl e7 1 7.i.d3 with a pleasant position) 1 5 . �d2 'lWxf4?! 1 6.i.xf6 'lWxf6 17. lLl e4 �f4 1 8 . lLl d6t rJde7 1 9.93 �xd6 20. �xd6 rJdxd6 2 1 . 'IW f2 with a decisive plus in Guseinov - Mamedov, Baku (U-20) 2002.) l 1 .fxg5 lLlg4 12 .i.e2 lLl ge5 1 3 . lLl f3 i.e7 14 . �hgl !?

Initially played by Ziatdinov in a couple of striking outings, this indirect defence of g5 ensures that Black will not win back his pawn so easily. Moreover, the considerable tension that arises in the position greatly restricts his options, and f7 is potentially very vulnerable.

14 . . . b5 ( 14 . . . �g8? 1 5 .i.g3! hxg5

16 .lLlxe5 lLlxe5 17.i.xe5! dxe5 1 8 .i.h5 heralds a nasty accident on f7. Ziatdinov - Barbero, Montpellier 1 994) 1 5 . �dfl ! . I t i s important to free up the dl square as a comfortable retreat for the knight. 1 5 . . . b4 16 .lLldl �b8 17 . rJdbl a5 1 8 .b3! . The key move. Since e3 is out of bounds due to the 'g5 situation' the knight is headed for c4 via b2. Black has tried:

b I ) 1 8 . . . hxg5 ( 1 8 . . . lLlg6!? 19 .93 lLl ce5 is met with the immediate 20.gxh6! although after 20 . . . lLlxf3 2 1 . �xf3 lLl e5! ? this still might be Black's best choice) 19 .i.xg5 lLlxf3 (maybe 1 9 . . . i.xg5 20. lLlxg5 a4 (Gofshtein) 2 1 . lLl b2 axb3 22.cxb3 but both f7 and d6 remain problems) 20.gxf3 i.xg5 2 1 . �xg5 �f6 was Acs - van der Wiel, Wijk aan Zee 'B' 2003, when 22. �fgl ! ? 'lWd4 23.i.d3! rJde7 24.f4 looks good for White according to Acs and Hazai.

b2) 1 8 . . . a4 1 9 .1Llb2 axb3 20.cxb3 �a5? ! . A mistake which nicely illustrates the problem of altering the delicate balance of force around g5 . (20 . . . hxg5 2 1 . i.xg5 i.xg5 22. lLlxg5 �a8 is better though either 23. lLl a4 or 23.i.b5 look quite promising) 2 1 .g6! fxg6 22.i.xe7 rJdxe7. This exchange is pretty disastrous for the black king. Not just d6 but also the dark squares on the f-file become very exposed. The speed of the denouement is still striking. 23 . �dl �c7 24. lLlc4 lLlxc4 25.i.xc4 lLla5 26.e5! d5 27.i.d3 �c3 28. 'lWf4 1-0. Karjakin - Nijboer, Wijk aan Zee 2003.


Back to the main game with 9 . . . b5 . lO.i.xf6

Clearly the most principled continuation. lO . . . gxf6

10 . . o'IW xf6 is inferior due to I l .e5 dxe5 1 2. tLl dxb5 �d8 13 . tLld6t !xd6 14.�xd6 exf4 when either 1 5 . tLl e4!? or 1 5 .!xa6 �xa6 1 6. tLl b5 �a7 17.tLlxa7 tLlxa7 18 .�d4! tLl c6 19 .�xg7 �f8 20.�d2! offer good chances.

l l .i>bl Wb6 12.tLJxc6 i.xc6 13.f5!? For a long while 13. � e1 has been the

most fashionable move here. The claim is that by floating the idea of tLl d5 into the equation, White forces his opponent to make some kind of concession. It has been a long haul for me towards the conclusion that, while Black may enter this system with high hopes of a more than usually active role for his king's bishop, in many cases the unpretentious e7 is not such a bad square and slightly misplacing the

queen in order to coax it there is not necessarily best.

Specifically Black's counterplay in the variation 13 . . . !e7 14.f5 �c5 1 5 .!d3 b4 1 6. tLl e2 a5(!) (likely to replace the older 1 6 . . . e5 which aims to restrict White's knight, but in fact offers it a choice of promising routes via g3 or c 1 ) 17.fxe6 fxe6 1 8 . tLl f4 �e5 19 . �f1 a4 20 . .tc4, Nijboer - Acs, Wijk aan Zee 2003, when by flicking in 20 . . . a3! it seems that White could be caused some embarrassment. Hence we shall try to expend no tempi on moving the queen, even if it is necessary to sacrifice a pawn in the process.

13 . . . b4 Highly committal. There is the danger

that if Black is not actually taking on e4 then the text may merely be driving his opponent's knight to a better square, offering White's bishop more promising options than d3 and rendering any future . . . 0-0-0 highly problematic. Rather a

1 30 The Classical

menu of drawbacks for any move to bear.

For all these reasons I am more inclined towards the solid 1 3 . . . 1Wc5! ? eg 14 .i.d3 h5 1 5 .1Wel ! ? (Only now. The tempo expended on . . . h5 means that Black can no longer get the level of counterplay described in Nijboer -Acs above. Waiting for . . . h5 to play this queen move seems like a fair rule of thumb.) 1 5 . . . 0-0-0 (If now 1 5 . . . i.e7 1 6. itJ e2 e5 17.h4!? (fixing the h5 pawn in order to target it) 17 . . . a5 I B . itJ g3 'it>d7 1 9 .i.e2 l::1agB 20.i.f3 a4 2 1 . itJxh5 a3 22.b4! 1Wc4 23.g4 with clear advantage Kosteniuk - Bu Xiangzhi, China-Russia, Shanghai Ouniors) 200 1 .) 1 6. l::1f1 i.h6!? 17.fxe6 fxe6 I B . l::1xf6 1We5 1 9.1Wf2 i.b7 ( 1 9 . . . i.e3? ! is risky due to 20.1Wxe3 1Wxf6 2 1 .a4! (2 1 .1Wb6 i.b7 22.i.xb5 axb5 23. itJxb5 l::1d7 is just a draw after 24. itJa7t etc) 21 . . . 1We7 22.axb5 axb5 23 . itJ xb5 with promising compensation) 20. l::1f7 l::1hfB 2 1 .a3 l::1xf7 22.1Wxf7 l::1fB 23.1W g6 i.e3 24.i.e2 and although Black has some activity for the pawn, there are still enough weaknesses for White to aim at too. Lastin - Palac, European Club Cup, Panormo 200 1 . 14.ltJe2 e5

This has been played on several occasions, and a sneak look back through my private files suggest that a couple of years ago I spent a while on this variation without so much as suspecting Lastin's superbly economical novelty on move 17 . Taking the pawn is also a serious option,

but White's compensation looks quite attractive enough to recommend the line, and indeed there may even be a decent choice after 14 . . . i.xe4 1 5 .fxe6 ( 1 5 . itJ g3!? is interesting but untested. Acs and Hazai consider 1 5 . . . i.b7 (or 1 5 . . . i.d5 1 6. itJ h5 0-0-0 17. itJ xf6 i.b7 I B .fxe6 fxe6 19 .1We2 d5 20.g3 The bishop pair is scant compensation for the weakness of e6.) 16.fxe6 fxe6 17.1We2 although after 17 . . . 'it>d7!? I would be happier if the pattern of White's further development was clearer) 1 5 . . . fxe6 and now there are two good choices:

a) 1 6. itJ f4 d5 (or 16 . . . b3!? 1 7.axb3 1Wxb3 I B .i.d3 i.xd3 1 9 .itJxd3 1Wc4 20. l::1he l e5 2 1 . itJ b4 - Acs/Hazai, again with definite compensation) 17.i.d3 0-0-0 l B . i.xe4 dxe4 1 9 .1W e2 l::1xd I t 20. l::1xd l e3 2 1 .1Wf3 i.d6 22. itJxe6 l::1 eB 23. itJd4 i.c7 24.g3 and White has a lightsquare advantage, although until he can round up the e3 pawn there will always be


tricks. Sadvakasov - Yermolinsky, Stratton Mountain 2000.

b) 1 6.�f4!? f5 ( 1 6 . . . d5 17.�xf6 Ek8 1 8 .tZld4 E:g8 19 .id3! - Acs/Hazai.) 17. tZl g3 �c6 1 8 . tZl xe4! �xe4 19 .�d2! was Acs - Duppel, Germany 2000, when after the best defence 1 9 . . . �c6!? 20.ie2 E:g8! Acs and Hazai give 2 1 . E:he1 but I also like the look of2 1 .if3!? d5 22. E:he 1 0-0-0 23. E:cl ! and the coming c4 break will cause a real headache, en passant notwithstanding. 15.�g3 h5 16.h4! YlYc5

Damljanovic's latest try was the pawn sacrifice 1 6 . . . ih6!? 17 . �xd6 E:d8 1 8 .�xd8t �xd8 1 9 . E:xd8t 'it>xd8 20. tZl xh5 ixe4 2 1 . tZlxf6 ixf5 22.ixa6 ie6 23.ie2 'it>e7 24. :Ei:fl ie3 25 .g3 id4 26. tZl e4 ih3 27. E:f6 ie6 with the bishop pair spearheading quite decent compensation. However, 1 7. � e2! ? looks very plausible here too. 17.YlYe2!

Another example of a novelty of 'why didn't I think of that' simplicity! Moving the queen immediately takes the sting out of defences based upon the activation of the bishop commencing . . . ih6, while White has nothing to fear from 17 . . . ib5?! since the exchange of lightsquared bishops is near the top of his menu of positional goals. 17 . • . a5 18.�xh5 i.e7 19.94 a4 20J:�h3!

Another excellent move with both defensive and aggressive designs.

20 • • • a3 21 .b3 1�k8 22.YlYg2!

Now Black is also helpless against ic4, almost always a powerful move in this variation if it can be effected safely. The speed with which White accumulates almost every advantage imaginable against a player of Kevin Spraggett's calibre is quite awesome and an excellent advert for Lastin's treatment. 22 • • • YlYb6 23.i.c4 i.b7 24.YlYe2 gg8 25.g5 fxg5 26.f6!

The rest is a massacre and requires little comment. A very powerful display, and I suspect we shall see a good deal more of these treatments without an early � e 1 in the future. 26 • • . i.f8 27.gg3 gxc4 28.bxc4 YlYc6 29.hxg5 i.a6 30.YlYd2 hc4 3 1 .YlYxb4 YlYxe4 32J�b3 YlYc6 33.gxa3 d5 34.�g7t �d7 35.ga7t �c8 36.YlYa5 i.c5 1-0

1 32 The Classical

Game 26 Dolmatov - Makarov Samara 2000

1 .e4 cS 2.ltJa ltJc6 3.d4 cxd4 4.ltJxd4 ltJf6 S.ltJc3 d6 6.J.g5 e6 7.'flYd2 J.e7

With the customary apologies to historical record I have fiddled with the move order. In fact Makarov chose 7 . . . a6 8.0-0-0 ltJxd4 9. \1!!fxd4 !e7, but it useful to invert this in order to deal conveniently with a number of minor lines commencing 7 . . . !e7 here. First though, this seems like a good moment to consider various other 7th move choices:

a} 7 . . . h6 does not impress. As we have seen, every Rauzer player should be ready to play the familiar structure with the doubled f-pawns under certain circumstances, but the enthusiasm for it embodied in this loss of tempo is regarded as slightly eccentric. White secures good play with something like an extra tempo over lines considered under Game 23 note c2 to 6 . . . \1!!fb6. In Z. Almasi -Damljanovic, Cacak 1996, White gave a convincing example of how to handle this structure after 8 .!xf6 gxf6 (8 . . . \1!!fxf6? 9 . ltJdb5) 9.0-0-0 a6 1 O.f4 !d7 1 1 . @b1 \1!!fb6 12 . ltJ b3 0-0-0 13 .!e2 @b8?! (Black should at least try to prevent White's next by 1 3 . . . h5, although 14 .h4! ? looks appealing then when the h5 pawn is distinctly vulnerable.) 14 .!h5! �h7 1 5 . �hf1 !c8 16.�e2 \1!!fc7 (if 16 . . . J.e7 then 1 7. ltJ d5! is strong - a common

motif in such positions) 17. ltJ d4 ltJxd4 1 8 . �xd4 !g7 1 9 . �c4 \1!!fe7 20.5 �hh8 2 1 . \1!!f d3 �hg8 22.a3 and all activity and life is being squeezed out of Black's game. A rosy version of how to handle this structure, but the f4-f5 plan has featured many times, and this model execution is a useful game to know, as well as a warning against 7 . . . h6.

b} 7 . . . \1!!fb6 poses questions for our repertoire built as it is around 9 .f4 rather than 9. ltJb3 in the main line of game 25 . Here White can again consider 8 .!xf6 likely to transpose to positions dealt with under the note to 6 . . . �b6 to game 23. However, as I said there, I am quite keen not to force these difficult positions upon the reader. So another alternative with Garry Kasparov's stamp of approval. 8 .0-0-0!? �xd4 (8 . . . a6 9 .ltJxc6!? bxc6 10 .!c4! !e7 l 1 .!b3 !b7 ( 1 1 . . . e5 12 .J.xf6 gxf6 13 .f4±) 1 2.f4 �d8 1 3 .5 e5 14.g4! h6 1 5 .!xf6 gxf6 1 6. @b1 was terrific for White in Vorobiov - Avrukh, Moscow 2002. The only mystery here is why there have been few imitators; the idea looks very fine.} 9 .�xd4 ltJxd4 1O . �xd4 a6

Of course, the simplification may not be to the taste of those who like a quick knock-out, but this is definitely one of the set of Rauzer endings in which I find it inconceivable that White's space advantage should count for nothing. In particular, bear in mind the weakness of b6.


b l ) 1 1 .!xf6!? gxf6 12 . � a4 !d7 1 3 .�b6 Eld8 14 . �xd7 Elxd7 1 5 .!e2 Elg8 1 6.g3 !h6t 1 7.cj;>b l cj;>e7 1 8 . Elhdl Elc8 1 9 .EI Id3 Elc5 is not terribly exciting, with undeniable drawish tendencies, but there are pretty well only two results in play which will appeal to those who like to play very safe. It is unlikely to appeal to Black! J. Fernandez Garcia - Estremera Panos, Spain (ch) 2002.

b2) 1 1 .f3 !d7 1 2 . � a4(!) was Kasparov's contribution which, as is customary, set the stage for subsequent tussles. Critical is probably 12 . . . d5 ( 12 . . . e5 1 3 .�b6! favours White, whose knight by one route or another will enjoy the excellent outpost on d5; while after 1 2 . . . !c6 1 3 . � b6 Eld8 14. � c4 !e7 1 5 . � a5! Elc8 1 6. Elb4! d5 17. �xc6 bxc6 1 8 . Elb7 � d7 19 .!d2! . White was richly rewarded for his creative manoeuvres in the influential Kasparov - Mchedlishvili, Bled (01) 2002) 1 3 .exd5 �xd5 14 .!c4

!xa4 1 5 .!xd5 exd5 16. Elxa4 f6 17.!e3 cj;>f7 18 .Eldl b5 1 9 . Elad4 !c5 20. El4d3 !xe3t 2 1 . Elxe3 Elhe8 22. Eldel Elxe3 23.Elxe3 Ele8 24.Elxe8 cj;>xe8 25. cj;>d2 cj;>d7 and Black held the pawn ending in Volokitin - Gershon, Bermuda 2003 .

It might be worth checking out 13 .!xf6!? though. After 1 3 . . . gxf6 14.exd5 White's coordination may suffer a little after 14 . . . e5 1 5 . Elh4, or 14 . . . !h6t 1 5 . cj;>b l e5 1 6. Elc4! b5 17. � b6, but whether that will amount to full value for Black's investment is rather more doubtful.

Now we return to the main game after 7 . . . !e7

8.0-0-0 a6 8 . . . �xd4 9 .'lWxd4 0-0 is a quite

legitimate transposition of moves leading to Game 27, while 9 . . . b5 1 0.f4 leads back to the main line here. 9.£4 �xd4 10.Y!Yxd4 10.£4 b5!�

1 34 The Classical

Almost unheard of just 1 0 years ago, this has already become another major Rauzer system, which has at any rate defied all the more brutal early attempts to kick it into touch. For this reason a solid system will be advocated here, heading for the doubled f-pawn structure but with due preparation. 1 1 . .ie2!?

This requires some explanation since immediately inflicting damage on the opponent's structure with 1 1 ..!xf6 is by far the more popular choice. In particular a word on two on the popular tries which in my view 'just miss' for White:

1 1 ..!xf6 gxf6! ( 1 1 . . . '!xf6? 1 2.e5 dxe5 13 .'�c5! ( 1 3 .'�e4! is also very strong) 1 3 . . . .!d7 14 . lLlxb5! meeting 14 . . . axb5 with the devastating 1 5 . 1hd7! while if 14 . . . .!e7 simply 1 5 . lLl c7t @f8 16.'�·xe5 is very strong according to Kasparov.) and now:

a) 1 2 . .!e2 Vlic7 1 3 .f5 Vlic5 14 .fxe6 fxe6 1 5 .Vlixc5 dxc5 16 . .!h5t @f8 17.e5

f5 1 8 .g4! was a line which put me off the black side of this variation for some time. White opens the position against the bishop pair to embarrass the black king and to prise open nice squares for his knight. However, with precise defence it seems that Black can neutralise the White initiative, and with precise handling the bishops can still be a major defensive asset. 1 8 . . . E:a7 1 9 .9xf5 exf5 2o. lLl d5 .!g5t 2 1 . @b l '!e6 22. E:hgl E:d7!? (also Curt Hansen's 22 . . . E:g8 23. lLl f4 .!c4! seems tough to refute since 24.b3 is well met with 24 . . . .!xf4!) 23. E:xg5 E:xd5 24. E:xf5t @e7 25. E:g5 E:g8 26Jhg8 E:xdl t 27 . .!xdl .!xg8 28 . .!g4 .!e6 29.'!f3 .!f7 and this ending should be a draw Grischuk - Grosar, Batumi (European-ch) 1 999.

b) 12.e5 d5 1 3 .@bl .ib7 14.f5 has been ultra-trendy lately and is certainly the fiercest weapon at White's disposal. However, after 14 . . . fxe5 1 5 .�xe5 .!f6 16 .Vlig3 �e7 17.fxe6 fxe6 1 8 . .!e2 h5


I do find myself disturbed at the large number of positional concessions that White is required to make. The position reminds me of the material of game 24, a true battle between strategic and dynamic features, and yet there it is somehow more believable that White's initiative is durable and represents genuine compensation. It is true that the clever break 1 9 .a4 is still troublesome for Black since 19 . . . b4 is met by 20. 'lJ e4! dxe4 2 1 .�xg6t Wfff7 22.�xf7t and White recoups the piece with interest. However, I am attracted to the deflecting 1 9 . . . h4!? of Kovalev -Supatashvili, Leon 200 1 . 20.Wffg6t Wfff7 2 1 . �d3 when 2 1 . . . 0-0!? was OK, and even 2 1 . . . bxa4 looks worth a look. I offer this so that the reader has a starting point for keeping pace with shifts in high fashion, but personally speaking I am fundamentally sceptical about White's position. 1 l . . . ib7 12.1a

What has this manoeuvre achieved? Well, the e-pawn is well defended and the e2 square cleared for the knight, both useful preparations for the familiar strategy of pushing f5 against doubled f-pawns, and also as we shall see, handy too if Black wishes to keep his structure intact. More subtle, but no less important, the whole plan with f5 seems far more effective with Black's bishop already committed to b7. Indeed it was this nuance that finally persuaded me that this set-up was worthy of the repertoire.

12 . . . �c7 Or 12 . . . Ek8 13 .1xf6 and now: a) 1 3 . . . i.xf6?! attempts to cross White's

plans. Black doesn't even lose a pawn, but as is often the case in the Sicilian, the move . . . b5, so useful in the middlegame, gives White a handy target for operations in the ending after 14 .Wffxd6 Wffxd6 1 5 . Ei:xd6 1xc3 1 6.bxc3 Ei:xc3 17. Ei:hdl 0-0 1 8 . Ei:d7! 1c8 ( 1 8 . . . 1c6 19 . Ei:c7 b4 20. Ei:d4 i.b5 2 1 . Ei:xc3 bxc3 22.a4 1f1 23. �dl Ei:c8 24. �el 1c4 25 .1e2 1xe2 26. �xe2 also left Black very overstretched in the rook ending in Chandler - Fernando, Santo Antonio 200 1 ) 1 9 . Ei:a7 Ei:e8 20. Ei:d4 �f8 2 1 . �b2 Ei:c5 22.e5 Ei:e7 23. Ei:d8t Ei:e8 24. Ei:d3! Ei:e7 25 .Ei:xe7 �xe7 26. Ei:c3! Ei:xc3 27. �xc3 �d7 28. �b4 �c7 29. �c5 and White's superb king is a decisive asset in the bishop ending. Malakhov - Blehm, Cappelle la Grande 2000.

b) 13 . . . gxf6! 14.f5 and now:

1 36 The Classical

b I ) 14 . . . EkS I s . mb l Vffc7 16 .Vffd2 hS(?!) My hunch is that the weakness of g6 outweighs the coverage of the hS and h6 squares. 17.fxe6 fxe6 I S . tLl e2 .!f8 19 . tLl f4 striking at the dual weaknesses on e6 and g6. Again, I am struck by how the b 7 bishop would be happier never to have moved. 1 9 . . . mf7 20.Vffe2 h4 2 1 . E1hfl .!e7 22.'!hSt E1cxhS 23. tLlxhS Vff c4 24. Vffxc4 bxc4 2S . tLlxf6! mg6 26.eS .!xg2 27. E1fe 1 dS 2S. E1d4± Kreiman -Mainka, Bad Zwesten 1 995.

b2) 14 . . . E1c4 ! ? , gaining a tempo to double on the c-file I S .Vffd3 Vffc7 1 6.fxe6 ( 1 6. tLl e2!?) 1 6 . . . fxe6 17. tLl e2 VffcS I S .mb l ! VffeS 1 9 .tLld4 hS (Again I am not convinced by this move. Shirov gives 1 9 .. .!cS! ? However, White will build with g3/ .!g2 etc and Black lacks counterplay.) 20. E1he l md7 2 1 .g3! E1gS? 22 . .!g2 E1g4?! 23. tLl£3 WfcS 24. tLld2!± E1a4 (24 . . . E1d4? 2S . tLl b3!) 2S . .!h3 E1gS?! 26. tLl£3 E1g7 Shirov - Pelletier, Elista (ol) 1 995, when 27.b3! E1aS 2S. tLl d4 E1g4 29 . .!xg4 hxg4 30.c3 - Shirov, would have been immediately decisive.

Of course these examples are not clearcut, but the simplicity of White's plan and the economy of force with which he can attack on the light squares suggest that this is a promising approach. The main game provides further promising evidence. 13.ixf6 gxf6 14.f5 h5

Yet again, this does not seem strictly necessary. However, the complexity of

Black's defensive task is rather reflected in the complete lack of agreement among commentators. There is in particular no consensus around whether Black should be looking to castle long, or whether his rook belongs on cS. If it was a check on hS Black feared, then I think the text was questionable. After 14 . . . E1cS I S .fxe6 fxe6 neither 16 .'!hSt md7 nor 16 . .!g4 mf7 17 .'!hSt mg7 I S .eS ( 1 S . .!g4!?=) IS . . . E1fdS!

G. Garcia - Smirin, New York 1 997 really convinces. However I S . Vff d2! ? looks interesting. White will follow up with mbl and tLl e2, while if I S . . . Wfc4?! 1 6. mbl b4 17. tLl e2 .!xe4 IS . .!xe4 Vffxe4 1 9 . tLl g3 Wfc6 20.fxe6 fxe6 2 1 . E1he 1 mf7 22 . Wf e2 dS 23 . tLl f5 gives a strong attack. The argument against 14 . . . E1cS might indeed be that . . . 0-0-0 was needed in reserve!

15.fxe6 I see no special objection to this

exchange, but White could also keep the tension, since after 1 5 . <;t> b 1 1& c5 the untried 1 6.1&xc5 !? dxc5 17.fxe6 fxe6 1 8 .i2:le2! h4 1 9 . i2:l f4 looks quite promising. IS ••• &OO 16.<bbl O-O-O�!

As we have seen, where to put the king is a riddle with no easy answer for Black in this line. However at this particular moment there is no impediment to White immediately compromising the king's new home, and we can give a fairly unambiguous assessment of this decision. 17.a4! dS

Looks a bit desperate, but stronger than the horrible 17 . . . bxa4 18 .1&xa4 when 1 8 . . . E!:hg8 1 9 .1&b3! 1&d7 20. i2:la4! -Gofshtein, illustrates one danger lurking! IS.adS b4 19.tLle2 <bbS�!

For better or worse, 19 . . . ixd5 had to be tried. White has a positional plus based on Black's light-square looseness both in the centre and on the queenside after 20.ixd5 E!:xd5 2 1 .1& e4! but he can struggle on. 20.'ilYe3 eS 21 .tLlg3! ics 22.'ilYb3 h4 23.tLlf5 ics 24.d6!

It is still all about White's light-square chances, but now they take on an added ferocity. White answers threat with counter threat, and the momentum never dies down! 24 ••• 'ilYb6 2S.aS 'ilYbS 26.'ilYf7! E!:d7 27.'ilYxf6 geS 2S.tLle7 ib7 29.ixb7 'ilYxb7 30.'ilYxeS .hd6 31.tLlc6t 1-0

Game 27 Iordachescu - Campos Moreno Linares 2000

1 .e4 cS 2.tLla tLlc6 3.d4 cxd4 4.tLlxd4 tLlf6 S.tLlc3 d6 6.ig5 e6 7.'ilYd2 te7 S.O-O-O O-O

After 8 . . . i2:l xd4 9. Wi xd4 a6 was covered in game 26, while 9 . . . 0-0 is also a quite legitimate move order, but since it is designed primarily to circumvent lines with 8 . . . 0-0 9 . i2:lb3, it causes no inconvenience to our proposed repertoire and White should just return to the main game with 10 .f4.

9.£4 tLlxd4 Black can also put the question to

White's bishop immediately with 9 . . . h6, hoping to find situations in which its unprotected status on h4 will become a real issue. In addition, whereas there is often a sacrificial option available to the attacker in the event of a later . . . h6, choosing this

1 3B The Classical

moment has the important virtue that the soundness of 1 0.h4?! is in real doubt (at least given that White is playing for a win) and therefore I am recommending that White accept the need for 10 .ih4. The drawback for Black is that there are also cases where this square proves a good deal more comfortable (see in particular (d) below) . Indeed the limitations of this piece's supposed vulnerability are immediately evident in the event of 10 . . . lthe4?! l 1 .ixe7 lLlxd2 12.ixdB lLlxfl 1 3 .lLlxc6 bxc6 14 .ie7 E1eB 1 5 . E1hxfl E1xe7 1 6. E1xd6 ib7 17.g3! c5 I B . E1fdl Unzicker - Stahlberg, Moscow 1 956, when although Black's weaknesses do not leave him 'lost' his defensive task is certainly unenviable.

After l O.ih4, Black does nonetheless have quite a wide choice:

a) 1 0 . . . Wlb6 has been pretty much condemned since White's powerful display in Leko - Kramnik, Belgrade

1 995 . In the structure which arises from the forcing sequence l 1 . lLl xc6! bxc6 1 2.e5 dxe5 1 3 .fxe5 lLl e4 14. lLlxe4 ixh4 1 5 .Wlf4!, Black's bishop pair is a decidedly mixed blessing, indeed it seems to me that he sorely misses his knights. White's kingside aspirations are a bit crude, but rather effective, and although Kramnik undoubtedly missed a chance to generate a bit more on the other wing, the theoretical verdict remains intact. 1 5 . . . ie7 16 .id3 E1bB 17.b3 WI c7?! (A bit cooperative. On grounds of both offence and defence 17 . . . Wlb4!? makes more sense. However, even then I prefer White's chances after I B . E1hfl a5 1 9 .'it>bl a4 20. E1f3 axb3 2 1 .cxb3 VNa3 22. E1g3 E1dB?! Wapner - Pavasovic, Bled 1 996 when with 23. E1fl ! White could have exerted extreme pressure. Instead, 22 . . . 'it>hB! makes more sense, but 23.VNg4! E1gB 24. E1fl is still promising. Black tends not to be really threatening to sacrifice on b3, provided the splendid knight on e4 stays put!) l B . 'it>bl c5 1 9 .1Llf6t 'it>hB 20.Wle4 g6 2 1 .h4 ib7 22. Wlf4 'it>g7 23. lLlg4 E1hB 24. E1dfl E1bfB 25. lLlf6 WldB 26.VNg3 h5 27.Wlg5 E1h7 2B. E1f4 with utter domination.

b) 10 . . . VNc7 has still more emphatically disappeared without trace. However, the 'refutation' is not just convincing, but also quite specific and hence worth being aware o£ White should play 1 1 . lLl db5 Wla5 12 .VNe l ! E1dB 13 .a3! E1d7 14 .ixf6 ixf6 ( l4 . . . gxf6 15 . E1d3 does not bear

thinking about!) 1 5 . � xd6 i.xc3 16 .vgxc3 vgxc3 17.bxc3 e5 I S .f5 <j;>fB 19 .i.e2 as in Yudasin - D. Gurevich, Beersheba 1 993, and since 1 9 . . . <j;>e7?? now loses to 20. �xcSt 1'!xcS 2 1 .f6t gxf6 22. 1'!xd7t <j;>xd7 23.i.g4t Black has particularly awkwardly placed pieces in addition to his pawn deficit.

c) 1 O . . . i.d7 1 1 . � f3! (directed against . . . �xd4 and . . . i.c6 with the hope of showing that Black's 1 0th move is fundamentally rather passive) 1 1 . . . 'Wa5 1 2. <j;>b1 1'!fdS 13 .i.d3 and now:

c 1 ) 13 . . . b5 14.g4! . This dangerous pawn storm is White's main strategy here, especially as 9 . . . h6 has created a useful target. 14 . . . b4 1 5 .i.xf6 i.xf6 16. � e2 e5 17.g5 hxg5 I S .fxg5 i.e6 1 9 . � c1 i.e7 20.h4 d5 2 1 . vg e2 1'!d6 22.h5 1'!adS 23.g6 gave White a very powerful attack in Tiviakov - Shmuter, St Petersburg 1 993. It is instructive that Black's execution of the standard central break with 20 . . . d5 here does nothing more than create a rather meaningless 'tension' in the centre, neither impeding White on the kingside nor enhancing Black's efforts on the other wing.

c2) 1 3 . . . 1'!acS !? 14. 1'!hgl (the immediate 14.g4 is also possible, but Black can try 14 . . . �xg4 1 5 .i.xe7 �xe7 1 6. 1'!hg1 1'!xc3! ? 1 7.vgxc3 ( 17 . 1'!xg4 1'!c5 I S . 'W g2 g6 doesn't really convince) 17 . . . vgxc3 l S .bxc3 � f6 which is quite hard to assess) 14 . . . e5 1 5 .g4! exf4 1 6.g5 hxg5 17. 1'!xg5 � e5

l S . 1'!dgl ! (the safe l S .�xe5 dxe5 1 9 . � d5! is also sufficient to suggest that White has won the opening battle, but the text is much more incisive) I S . . . �xf3 1 9 . 1'!xg7t <j;>fB 20.'Wxf4 �xgl 2 1 . i.xf6 i.xf6 22.vgxf6 i.e6 23. 1'!h7 <j;>eS 24.i.b5t 1'!c6 25 . 1'!h5 d5 26.exd5 1'!xd5 27. 1'!xd5 i.xd5 2S.vge5t <j;>fB 29.vghSt <j;>e7 30. �xd5t <j;>d6 3 1 . 'Wh6t <j;>c5 32.i.xc6 vge1 t 33.vgc1 vgxc1t 34. <j;>xc1 bxc6 35 . � f4 � f3 36.h3 <j;>d4 and Black was active enough to hold the ending in Ivanchuk - Cu. Hansen, Skanderborg 2003. However 25 .'Wg5 !? d5 26.exd5 1'!xd5 27.i.xc6t bxc6 2S. 'Wxg1 retains material and positional plusses, while restricting any counterplay.

d) 1 0 . . . �xd4 1 1 .'Wxd4 'Wa5 has, by contrast with the moves considered above, recently enjoyed a real renaissance. In the event of 12 .i.c4! it quickly becomes clear that the move can only really be deployed in conjunction with a quick . . . e5, which

140 The Classical

represents a certain positional concession for which Black seeks solace in the bishop pair and some activity. The fate of 12 . . . i.d7?, in the style of the main game, emphasises the problem that the white bishop enjoys a far more secure existence on h4 - 1 3 .e5! wins material immediately. 1 2 . . . !'!d8 1 3. !'!hfl ! is also problematic since White can open the f-file and cause trouble there. Hence 1 2 . . . e5 1 3 .fxe5 dxe5 14 .Wfd3 i.g4 1 5 . !'!dfl ( 1 5 .i.xf6 is similar, but seems to me to allow Black more options - White can usually throw in this move when required) and now:

d 1 ) 1 5 . . . i.e6 (the bishop on g4 can be vulnerable. For example 1 5 . . . !'!ad8?! 16 .Wfg3 i.e6 17. !'!xf6 i.xc4 1 8 . !,!xh6±) 1 6.i.xf6!? i.xf6 17 .@bl Wfc5 1 8 .i.b3 @h8 1 9 .h4 i.e7 20.Wfe2 Wfc8 2 1 . i.xe6 fxe6 22. !,!xf8t WfxfS 23.Wfb5 Wfc8 24 . .!tJ e2 Wfc7 25 . .!tJc 1 ! !'!c8 26.'1We2 with pressure Ponomariov - Avrukh, Panormo 200 1 .

d2) 1 5 . . . Wfc5 16 .i.b3 @h8 17 .i.xf6 i.xf6 1 8 . @b l !'!ad8 1 9 .Wfg3 i.e6 20 . .!tJd5 i.xd5 2 1 . i.xd5 is very typical of the small but pleasant advantage White can expect here. He has the far superior bishop, and chances to activate his rooks, both on the f-file and swinging along the third rank. Kobalija - Gershon, Kharkov 2002.

e) 10 . . . e5!? nonetheless remains by far the most important of these 'side-lines' .

1 1 . .!tJ f5 i.xf5 12.exf5 exf4! 1 3 .@b l d5! 14 .i.xf6 i.xf6 1 5 . .!tJxd5 i.e5 i s a fairly well established sequence at the end of which we reach a position in which the key factor is the presence of opposite coloured bishops and with it a degree of dominance by the respective sides over 'their colour domains' . There are two dangers to be aware of - that Black will generate rapid counterplay based upon his excellent bishop on e5, and that the advanced doubled f-pawn can be


consolidated and prove a bit of a thorn in the side. However, I am confident that White's light square play should be the more durable if he carefully exerts pressure not just on the kingside but in the centre too. 16.c3!?

This quiet but far from innocuous continuation is likely to increase in popularity now that the 'main line' 16 . .!c4 has recently suffered a slight dent after 1 6 . . . b5 17 . .!b3 a5 l S .a3 a4 1 9 . .!a2 b4! 20. tihb4 'lWfG 2 1 . lLld5 'lWxf5 22.'lWd3 'lWxd3 23. Eixd3 g5 ! and it looks as though, in spite of White's passed pawn and his opponent's doubled pawn, the black pawns might be the more mobile. Tiviakov - Solak, European (ch) Istanbul, 2003.) Black has tested various moves here:

e 1 ) 16 . . Jl:bS 17.'lWc2 b5 l S . .!e2 EieS 1 9 .'!f3 'lWcs 20. Eihe 1 fG 2 1 .'lWe4 and the light square bind is in full flow.

Black tried 2 1 . . . b4 but was clearly worse after 22.cxb4 'lWb7 23. lLlxfGt .!xfG 24.'lWc4t \t>h7 25. EixeS EixeS 26 . .ixc6 'IW cS 27. 'IW c2 in Lutz - Lugovoi, Vienna 1 996.

e2) 16 . . . 'lWg5!? 1 7.'!d3 EiadS lS . .!e4 Eid6 19 .'lWc2 'lWdS 20.g3!? fxg3 2 1 .hxg3 lLl e7 22. lLl e3 :!=i:xd1t 23. :!=i:xd1 'lWc7 24. lLlg4 and White has enduring chances on the kingside which deny Black full equality. Simacek - Furman, Olomouc 2003.

e3) 16 . . . �e7 17 .'!c4!? ( 17. lLl xe7t 'lWxe7 l S . .!c4 might also yield a slight edge, but the text is richer in ideas.) 17 . . . lLl xfS (Critical, otherwise White's last would be an unambiguous improvement. 1 7 . . . lLlxd5 l S .'lWxd5 'lWfG 19 . Eihe 1 EiadS 20.'lWe4 gives a clear plus according to Kobalija, although I S . . . 'lWc7! ? might restrict White's initiative a bit. The fact that Black is still slightly uncomfortable as the position simplifies, largely due to light square targets on b7 and fl, is a reassuring sign for White.) I S . 'IW e2 :!=i:eS (Alternatives are no panacea either. l S . . . .id6 1 9 .1Llxf4 looks innocuous enough, but the Black pieces are actually quite loose, as is the fl pawn. For example 1 9 . . . EieS? loses to 20 . .ixf7t and 1 9 . . . EicS to 20. lLl e6! . 19 . . . \t> hS holds but looks unappetising. If l S . . . 'lWbS, Kobalija gives 19 . Eihfl , but 19 .93! ? played already in Mark Tseitlin - Kveinys, Gdynia 1 9S9 is effective against 1 9 . . . lLl e3 20.gxf4 lLlxdl 2 1 .fxe5 while 19 . . . fxg3 20 . .!d3! is exceptionally

142 The Classical

awkward!) 1 9 . .tb5! El:e6 20. ltJxf4 El:d6 2 1 . ltJ d3! . The key manoeuvre 2 1 . . . .tf6 22. ltJ c5! with a lasting and powerful plus on the light squares. Kobalija - Zaitsev, Russia 1 995. IO.Wxd4 Wa5

There are not a million ways to develop in this position and the text, preparing either . . . e5, or . . . .td7 by tactical means exploiting the pin along the 4th rank, is by far the most popular. As usual, 1 0 . . . a6 represents an alternative, although for some reason less popular at this juncture. 1 1 . � b 1 still makes sense to me, and after 1 1 . . .b5 White has the shot 12 . .txf6! .txf6 13 .e5 dxe5 14 .VNe4 VNb6 1 5 .VNxaS .tb7 1 6. El:d6! VNc7 17 .VNa7 VNxd6 l s .VNxb7 b4 ( 1 S . . . exf4!? 1 9 . .td3 .txc3 20.bxc3 would at least be less fun for White to play, although I do not believe that Black can claim full compensation.) 1 9. 1tJ e2 exf4 20.VNe4 El:dS 2 1 . ltJc l VNd1 22.g3 El:d4 23 . .td3 VNh5 Wells - A. Kiss, Hungary 1 996, and now 24.VNe2! would have been an eminently sensible way to consolidate. l l .i.c4

This might be seen by some as the 'old main line' , but in my opinion there is little doubt that it represents objectively the best try for an advantage - certainly when defending the Black side it is the line I have the most respect for. In addition it does have the serious practical virtue of being rather forcing, and results in a degree of simplification which retains

a decent level of tension while at the same time somewhat restricting Black's winning chances. Some commentators refer to it as leading to a boring endgame. Check this out - it is often the Black player annotating and trying to put his future opponents off the move!

1 l . . . J.d7 Black also has the sharp 1 1 . . . b5! ? at his

disposal, but I like 12 . .txb5 El:bS 1 3.a4!? (Securing the bishop in preparation for a quick e5) 1 3 . . . h6 ( 1 3 . . . a6 14.e5! dxe5 1 5 .fXe5 axb5 1 6.exf6 bxa4 17. ltJ e4! e5 I S . VN d5 is very strong, especially as after I S . . . VNxd5 1 9 .fXe7! VNa5 20.exfSVNt �xfS 2 1 . El:dSt VNxdS 22 . .txdS .tb7 23 . .tc7 El:cS 24 . .td6t wins material) 14 . .th4 e5 (Here too 14 . . . a6 fails to solve the problems since 1 5 .e5 dxe5 1 6.fXe5 axb5 17.exf6 gxf6 l S . .txf6 .txf6 1 9 .VNxf6 bxa4 20. El:d3 VNg5t 2 1 . VNxg5t hxg5 22. ltJxa4 leaves Black with little to show for the pawn.) 1 5 .fXe5 dxe5


1 6.'�'xe5 !e6 17 . lLId5 lLIxd5 l S .exd5 !xh4 1 9.dxe6 !g5t 20. @b1 !f6 2 1 .exf7t @hS 22.'�·d6± Campora -Moreno Tejera, Linares 2003. 12.e5 dxe5 13.fxe5 J.c6!

A good square for this piece, and a tactical solution that underlies Black's entire strategy. White's best is simply to retreat and secure the bishop pair. 14.i.d2! c!lJd7 15.c!lJd5 �d8

The trap 1 5 . . . 'Wc5! ? 1 6.!b4? ( l6 .lLIxe7t!) 16 . . . exd5 17.!xc5 !xc5 etc netting three pieces for the queen remains fairly victimless, but is worth being aware of! 16.c!lJxe7t �xe7 17J�he1

This may seem rather far into the game to be a describing the diagram position as a 'crucial starting point for further theory' , but there is some recompense for this in that the forcing nature and the strong underlying logic of the play do serve as a useful 'aide

memo ire' . In positional terms, there is also a fascinating and delicate balance of factors at work. White enjoys the bishop pair, more space and in many cases, whilst he might also reasonably aspire to attacking chances on the kingside, he will not be averse to an exchange of queens - Black's lack of room may even paradoxically be emphasised in such cases. Black has a solid position and hopes of either play on the queenside, and/or the chance to target the e5 pawn. Combating the bishop pair, there are also prospects of the knight and bishop working together to obtain chances on the light squares. 17 . . JUd8

Two others also deserve consideration:

a) 17 . . J HcS l S . !fl ! ? a5 ! l 9 .h4 b5 20.!g5 'Wc5 2 1 .'Wxc5 lLIxc5 22. !e7! with an edge for White in Lahno -Kotronias, Hastings 2003/4.

b) 17 . . . lLI b6 l S .!fl ! (This time l S . !d3? ! is well met by l S . . . E: fdS 1 9 .'Wg4 E:xd3 20.cxd3 !a4 2 1 . !h6 f5 ! 22.exf6 'Wxf6 23.!e3 E:cst 24. @ b 1 !c2t 2 5 . @a 1 .!xd1 which tends to be given as '= ' , but in fact the excellent black knight coming to d5 will give him at least an edge.) I S . . . E:fdS 1 9 . 'Wg4 and now:

b 1 ) 1 9 . . . a6? ! Weakens the knight on b6, and interestingly permits 20 . .!d3!? (threatening !g5) , with the nice point that after 20 . . . E:xd3 2 1 .cxd3 .!a4

144 The Classical

22 . .!h6 f5 23.exf6 �xf6 24 . .!xg7 �xg7 25 .�xe6t is very strong.

b2) 19 . . . �c5 looks a better try. However, it is here that White's rosy prospects in some of the endings come to the fore. 20 . .!h6 is possible, but I much prefer 20. �b4! and now:

b2 1 ) 20 . . . �xb4 (20 . . . �gl ? ! has a quirky appeal, but 2 1 . � f4! It:l d5 (2 1 . . . gd4 22 . .!e3!) 22.�h4! �c5 23.'!d3 g6 24 . .!g5 is clearly unsatisfactory for Black) 2 1 . .!xb4 It:l d5 22.'!d2!? b5 23 . .!g5 gdcB 24. gd4 a5 25 .c3 a4 26.a3! (that's about it for Black's queenside play. Now switch to the other side) 26 . . . h6 27 . .!d2 It:l b6 2B. gg4 h5 29. gh4 g6 30 . .!g5 It:l d7 3 1 .g4! with very strong pressure .!f3 32.gxh5 .!xh5 33 . .!xb5 It:l c5 34 . .!f6 'it>fB 35 .'!e2 'it>eB 36. gdl 1-0 Gallagher - Baumegger, Dresden (zt) 1 99B.

b22) 20 . . . �f2 2 1 .�f4! �c5 22.'!b4 gxdl t 23. gxdl �gl 24.c4 It:l d7

25 . .!d6 It:l fB 26.h3 It:l g6 27.�g3 a5 2B . 'it>b l a4 29. gel h5 30.�e3 Efimenko - Muir, Hastings Challengers 2003, and again the ending will favour White. For me personally the understanding that the bishop pair tends to be the number one factor in these endings has led to a degree of disillusion with 17 . . . lt:l b6, since the Black queen cannot flee the exchange indefinitely. 18.YlYg4 lt:lfS 19 . .id3!

Note 'b' above should not blur the fact that other things being equal this is the most threatening posting for this piece. The exchange sacrifice above was so potent precisely because the black knight on b6 was well poised for counterplay. Great respect though there should be for the beast on fB, its perspective is not primarily aggressive, and as we shall see, the coming exchange sacrifice will also be dominated by defensive considerations. 19 . . . gxd3

A familiar exchange sacrifice to nullify White's attacking prospects and his queenside majority, while also enhancing control of the light squares. When I last wrote about this more than six years ago, I stressed the fact that the play often reached an ending in which Black was made to suffer with nothing more than a half point to even aspire to. However, just before we went to press, Black was just waking up to the fact that his knight is a great piece and retaining it is more important even than ensuring that he

Experts vs. the Sicilian 14S

gets a pawn into the kitty. This has undoubtedly enriched the counterplay, although I am still generally enthusiastic for White.

The exchange sacrifice is not the only way to play. However, slow moves enable White to build up a serious initiative. 1 9 . . . Wl' c7?! , for example, was met with 20 . .tb4 idS 2 1 . @b1 !:1ac8 22.id6 Wl'c6 23.b3! bS 24. !:1e3 lLJ g6 2S .h4 f5 26.Wl'gS !:1d7 27.hS lLJf8 28. !:1d2 !:1f7 29. !:1f2! h6 30.Wl'f4 lLJh7?! 3 1 .g4± in Tatai - Epishin, Reggio Emilia 1 994. An excellent example of controlled aggression, keeping a tight grip on the position.

Personally, as Black I would still tend to favour the risky but apparently playable 1 9 . . . !:1dS!? The idea is clear - the rook exerts pressure on eS and also prepares to put further major pieces on the d-file. The drawback is equally apparent - the rook like the knight before it is vulnerable to attack from White's c-pawn and even in some danger of entrapment. After 20.ib4 Wl'd8 (But not 20 . . . Wl'c7 2 1 .c4! !:1d7 (2 1 . . .ibS? 22. @b1 ±) 22.id6 Wl'aS 23. @b1 lLJ g6 24.h4! f5 2S .Wl'g3 Wl'd8 26. !:1hl ! (26.hS Wl'h4!) 26 . . . lLJh8 27.cS! - cementing the bishop on d6, and activating its partner - 27 . . . bS 28.ic2! as?! 29. Wl'b3 White enjoyed a crushing position in Z. Almasi - Hracek, Germany 1 997. It is useful to have an illustration of what White can achieve when all the circumstances are favourable.) 2 1 . ib6 lLJg6

Black has two useful sources of counterplay - . . . Wl' as forking pawns, and . . . Wl'h4 commencing the task of probing White's kingside pawns - and one vital resource . . . f5. The latter can be critical. 22.c4?! would all but win here were it not for 22 . . . f5! when after 23.Wl'g3 the rook can nestle on d4 enjoying both security and swing potential. 22.g3 fails to impress either. After 22 . . . Wl'aS 23.ixg6 (23 . @b 1 ? lLJxeS) 23 . . . hxg6 24. @bl !:1c8 White is a long way from drumming up anything scary on the kingside. In general the exchange on g6 is only recommendable if a quick h4-hS follows. The best move from the diagram position therefore seems to be 22. @bl ! .

Now of course 22 . . . Wl'aS?! may be simply met with 23. Wl' g3 and it is instructive quite how difficult it is to strengthen Black's position. Moreover, the valuable role which the queen should be playing stopping h4 will soon become

146 The Classical

apparent - much as in the Almasi -Hracek note above to 20 . . . Vlic7?! .

Much more serious then i s 22 . . . Vlih4!? Then after 23. V!ixh4 lLlxh4 24.g3 lLlf3 25Jl:e3! (this position has as yet only been reached in Shomoev - Kharlov, ACP blitz (!) prelim 2004, and after 25J'!:e2 !b5! Black was fine) 25 . . . f5! ? (25 . . . lLlxh2 26.c4 virtually obliges 26 . . . gxd3 27. g 1xd3, which should be a reasonable version of the exchange sacrifice endings from White's standpoint. 25 . . . h5 26.h3! also leaves Black a bit stuck for what to do next) 26. gxf3 gxd6 27.exd6 �xf3 2B.gf1 and White has time to support his pawn with c4-c5 and should stand somewhat better in the ending. 20.0:d3 Y!Yd7 21 .ib4

21 . . . fllg6! Of course! Keeping the knight is now

almost always preferred to the rather depressing endings arising from 2 1 . . . Vli d5 22.�xf8! (the point. With this knight

gone Black's hopes are almost purely defensive) 22 . . . gxf8 23. @bl Vlixg2 24. V!ixg2 !xg2 25 .gc1 �c6 26. @c2!? f5 27.exf6 gxf6 2B. @c3!? 29. gf1 gh6 30. gg1 gg6 (30 . . . gh5 3 1 .h4!±) 3 1 . @d4 g4 32.b4 a6 33.a4 h5 (33 . . . !xa4 34. gcBt @f7 35 .gc7t @f6 36. gxb7±) 34. gxc6!? (not strictly necessary, but returning the exchange does clarify White's task) 34 . . . bxc6 35 . @c5 e5 36. @b6 gd6 37. @xa6 gxd3 3B . @b6 gb3 39. gal ! and basically the a-pawn runs home. Marciano -Lugovoi, St Petersburg vs. Paris, 1 996. Of course, this is a necessarily brief survey, but time and again Black's problem has been that undertaking anything active merely assists White in exchanging rooks! 22.J.d6 5!? 23.Y!Ye2!

An important improvement over 23. Vli g5 . Quite simply the queen is better placed for dealing with any counterplay Black may muster on the queenside. 23 . . . fllh8!?


Just how well prepared White is to combat standard light-square efforts is illustrated by the terrible tangle in which Black caught himself in Stefansson -Gershon, Bermuda 200 1 after 23 . . . .id5 24. 'i!7bl �c6 25 .�d2 .ixg2? 26J 'kl 1l;Yd5 27. Ek5 �f3 28.�gl .ifl 29. �g3 1l;YhI 30.�c1 ! ltJ f8 3 1 . �c7 'i!7h8 32. 'i!7al 1l;Yxh2 33. �c8 1-0.

So Black gives priority to evicting the annoying bishop on d6. Sensible, but the powerful knight he tried so hard to keep is otherwise less impressive on 0. 24J:�d2 �f7 25J�c2!?

White doesn't fear 25 . . . ltJxd6 26.exd6 .id5 27. � e5! with a decisive invasion on the 7th rank on its way. 25 . . J�e8 26.i.c5 Wd5 27.b3 a6?

A serious mistake. Better defensive chances were offered by 27 . . . .ib5 28. �c3 �c8 29 . .ib4 �xc3t 30 . .ixc3 .ixd3 3 1 . 1l;Yd2 .ie4 32.1l;Yxd5 .ixd5 33. �e2

- Gofshtein, although Black is a long way from securing the draw. Now, having evicted the bishop from one excellent outpost, Black offers it others. 28 . .ib6! Wh5 29.i.c5 gd8 30.d4 Wa5 31 .�b2 gd7 32J�ecl Wd8 33.Wc4 gd5 34.Wh4

It is clear that compared with the note above, White's coordination is quite admirable. All he needs is a plan to make further progress, and his excellent 36th move reveals this. The technical phase is a bit beyond our scope, but the Moldovan's handling is very sound. 33 . . . Wd7 35.a4 �g5 36 . .ia7! h6 37.Wb6 �h7 38.gxc6 bxc6 39.Wxa6 �e4 4OJ�xc6 gxd4 41 ..hd4 Wxd4t 42.�a2 Wd2t?! 43.�a3 Wxg2 44.Wc4 Wgl 45.Wxe6 Walt 46.�b4 Wd4t 47.�b5 Wd3t 48.�b4 Wd4t 49.gc4 Wd2t 50.�a3 Wfl 51.gc6 1-0

148 The Kan and Taimanov

Kan and Taimanov

- By Sune Berg Hansen

The Kan and Taimanov are both very sound and flexible Sicilian systems. I think that with good preparation it is possible to gain an advantage against both of these systems, but after some time surely a proper antidote will be found for Black and in the long run I do not think that either system can be refuted. They are popular at the highest level and players such as Anand and Kramnik use them occasionally. If White finds a good (or winning) idea in one of the sharpest lines he can be sure that he will only be able to use it once. Then all serious players will have fed it to their computer program (Fritz or Junior) and found an improvement. If they cannot find an improvement they will switch to another line. So in the computer age

one improvement usually means only one point. And it can take a week to find it!

The approach I will take in this chapter is a bit different. Instead of going for a refutation I will try to explain a system that is positionally well founded yet still active and aggressive. Therefore it can be part of the reader's repertoire for a long time.

In general the idea behind the chosen systems means placing the pieces on active squares and finishing development very quickly. White will attack with the pieces when possible. One of the reasons for this is to keep the risks involved to a minimum - when White throws the g- and f-pawns at Black's king he risks running into a nasty counterattack, and the endings also tend to favour Black. So the aim of this repertoire is what I will call controlled aggression: White will try to develop an initiative without taking undue risks. White's main ideas include: attacking Black's king (as usual in the Sicilian) and - more distinctively for the Kan and Taimanov - punishing b7-b5 if possible. I cannot guarantee a White advantage in every variation. But what I can promise is that you will usually know what you are doing, and why you are doing it! My ambition is that there will be no dead ends - there will always be play left in the suggested positions!

Let us start with a short introduction. First you should know that 1.e4 c5 2.�f3 e6 3.d4 cxd4 4.�xd4 a6 is the Kan


vanatlon (also called the Paulsen) , and l.e4 c5 2.c!LJa c!LJc6 (or 2 . . . e6) 3.d4 cx:d4 4.c!LJxd4 e6 is the Taimanov variation.

The Taimanov and Kan variations enjoy a special place and reputation in the Sicilian. These systems are not so easily understood as, say, the Najdorf, Sveshnikov or Dragon. One of the reasons for this is that they overlap each other (if Black goes c!LJ c6 in the Kan the game often transposes to the Taimanov) and they can also overlap with the Scheveningen and therefore some understanding of the Scheveningen is also necessary to avoid being 'move ordered' . Move orders are one of the biggest problems in the flexible Taimanov and Kan lines. I will begin by explaining some of the things that characterize the Kan/Taimanov. The big difference compared to the Scheveningen (1 .e4 c5 2.c!LJa e6 3.d4 cx:d4 4.c!LJxd4 c!LJf6 5.c!LJc3 d6) is, of course, that Black has not played d7 -d6 yet (and might not do so at all) or c!LJ f6, but he has the option almost every move! This is very important for understanding White's different lines against the Taimanov and Kan: Black is under no circumstances to be allowed to play d7 -d6 with a favourable transposition to the Scheveningen, or to a line that is not part of the repertoire recommended in this book.

Hence the first principle for playing against the Taimanov/Kan: Always remember Black can play d7 -d6 (and to f6) 'all the time' , and you should never

allow him to get a good Scheveningen! A lot of tricky players use Kan or Taimanov move orders to avoid lines like the .ig5-Najdorf, the English Attack, the Keres Attack and lines with .ic4 - this should not be achieved without paying a price.

Most of the variations I recommend against both the Taimanov and Kan are based on putting the white squared bishop on the active square d3. Before we start the survey of the proposed repertoire I will briefly mention White's other options. The reason is to avoid leaving the reader in the dark if he or she for some reason does not like the suggested repertoire.

Differences between the Taimanov and Kan

In the Taimanov Black has the knight on c6 from the beginning, this means that White has to keep the d4 knight guarded either by the queen from dl or by the bishop from e3. In the Kan Black often keeps the knight at b8, and most Kan players prefer a set-up with the knight going to d7 instead of c6.

Maroczy set-up Taimanov: l .e4 c5 2 . tOf3 e6 3.d4

cxd4 4. tOxd4 tO c6 5 .tOb5 d6 6.c4 tO f6 7. tO l c3 a6 8 . tO a3

Kan: l .e4 c5 2 . tO f3 e6 3.d4 cxd4 4. tOxd4 a6 5 .c4

White has the option of playing a set-up with pawns at c4 and e4. Some will argue

1 50 The Kan and Taimanov

that this is the 'punishment' for adopting the Kan and Taimanov instead of say the Najdorf, Classical, Scheveningen or Dragon variations where White forces lLlc3 before c4 by attacking e4 early. This line might give very good results for experts, but I do not think it fits very well with the rest of the repertoire suggested in this book. The lines are very positional in nature and I will recommend an attacking set-up instead. The c4-e4 system is better against the Kan than against the Taimanov where White has to put his knight at a3 in order to set up the Maroczy centre.

Recommended for positional 'system' players who enjoy a space advantage.

g3-systems Taimanov: l .e4 c5 2. lLlf3 lLlc6 3.d4

cxd4 4. tLlxd4 e6 5 .tLlc3 VNc7 (or 5 . . . a6 6.g3) 6.g3

Kan: l .e4 c5 2. tLlf3 e6 3.d4 cxd4 4. tLlxd4 a6 5 .lLlc3 VfIc7 (5 . . . b5 6.g3) 6.g3

Another option for White is to fianchetto his white squared bishop at g2. In general this is not a very critical approach to the open Sicilian. The exception is the Taimanov where the g3 variation ranks among the critical responses. The experts are Adams and Tiviakov. Against the Kan this line is pretty harmless.

Recommended for patient players with a solid style and who like a heavyweight positional battle.

1e2-Systems Taimanov: l .e4 c5 2. tLlf3 e6 3.d4 cxd4

4 .tLlxd4 tLlc6 5 .tLlc3 VNc7 6 . .te2 a6 7.0-o tLlf6 and now 8. 'it>hl or 8 . .te3

Kan: l .e4 c5 2. tLlf3 e6 3.d4 cxd4 4 .tLlxd4 a6 5 .tLlc3 (or 5 . .te2) 5 . . . VfIc7 6 . .te2

This is very popular and critical line against the Taimanov and a tricky sideline against the Kan. The biggest problem for White is that he has to be ready to accept a classical Scheveningen, which is not part of our recommended repertoire. At the moment I regard the .te2, .te3 line as White's most critical response to the Taimanov (this is what I fear the most when I play Black) . Against the Kan this does not make so much sense.

Recommended for principled players who have the time to keep up with current opening theory (the verdict swings between an advantage for White and 'unclear' almost every week) . Kasparov and Shirov play this successfully as White against the Taimanov. I think it is mostly recommended for players who also use .te2 against the Najdorf and Scheveningen.

English attack (J.e3 and £3) Taimanov: l .e4 c5 2 .tLlf3 e6 3.d4

cxd4 4. tLlxd4 tLlc6 5 .tLlc3 VNc7 6 . .te3 a6 7.VNd2

Kan: l .e4 c5 2. tLlf3 e6 3.d4 cxd4 4. tLlxd4 a6 5 .tLlc3 VNc7 (or 5 . . . b5 6 . .te3 .tb7 7.f3) 6 . .te3 tLlf6 7.f3


This is all the rage against the Taimanov at the time of writing. Against the Kan this does not make much sense as Black can go V!ic7, ib4 and d5 or b4 and d5 in one go. Again Black has the option of playing d7 -d6 transposing to a line that is not part of our recommended repertoire against the Najdorf/Classical.

This line is only recommended for players with plenty of time for opening studies. The verdict changes on a daily basis and we are still far from a 'stable' situation. This might be a good line for young aggressive (professional) players, but only against the Taimanov.

f4-systems White is currently doing very badly

with f4-systems in all lines of the Sicilian and this is therefore not recommended for anyone.

There are also some lines that are particular to the Taimanov and Kan move orders.

Special lines in the Taimanov White has the option of switching to

the Pelikan/Sveshnikov kind of centre with l .e4 c5 2 . tLl f3 e6 3.d4 cxd4 4. tLlxd4 tLl c6 5 . tLl b5 d6 6.,if4 e5 7.ie3. A line favoured by Leko.

If Black dislikes this he might use the move order l .e4 c5 2 . tLlf3 tLl c6 3.d4 cxd4 4. tLlxd4 'Wfc7 instead. This line does not promise much, but positional players with a technical style usually does not like to play with e5 instead of the more flexible (and elegant) positions with the pawn on e6.

Special lines in the Kan Black's set-up is flexible in the Kan,

but this also means there is no immediate pressure against White's centre (no knight on c6 or f6) . Therefore White can also stay flexible with 1 .e4 c5 2 . tLl f3 e6 3.d4 cxd4 4. tLlxd4 a6 5 .,id3 keeping the options of playing c2-c4 or developing the knight at


d2 instead of c3 open. This is regarded as the mainline and represents a whole different branch of the Sicilian (there is some overlap if White put the knight on c3 early) . This line is very complex and Black has no fewer than 10 different respectable set-ups! Anand and Adams are among the experts in this line.

Recommended for players with plenty of time for studying theory. This line can lead to very different kind of positions, and therefore White must have a universal style to be successful with this line.

The above characteristics are the opinion of the author and are not by any means certified truths.

Part I: 1he Kan

Against the Kan I recommend a system based on playing l .e4 c5 2. � f3 e6 3.d4 cxd4 4 .�xd4 a6 5 .�c3 followed by 6.Ad3. White has two standard ways to play these positions. The first consists of

playing moves like 0-0, WI e2, Ad2, gae l , f4, m h 1 and then either transfer a rook to h3 or go e4-e5 aiming for a kingside attack. The other way to generate active play is to start an assault on the queenside with a2-a4 (After Black's b5 . ) . This will usually be answered by b5-b4 after which White retreats the knight, and attacks b4 with c2-c3 leading to the opening of the c-file (Which is usually advantageous to White.) . White will get active play on the queenside and in the centre.

Overview of part I Game 28-30 covers the variation with

5 . . . Wlc7 which is the old mainline in the Kan. In Game 3 1 -34 the fashionable 5 . . . b5 i s covered. We begin with the game that has put the old Kan under a cloud.

Game 28 Acs - Fancsy Zalakaros 1 997

1 .e4 c5 2.�f3 e6 3.d4 cxd4 4.�xd4 a6 5.�c3

If Black tries 5 . . . d6 here, to sidestep the Keres Attack, White's best move is probably 6.g4! anyway (see page 277) . 5 .. . Wc7 6.i.d3 �f6 7.0-0 !c5

This is one of Black's main ideas. Before the bishop goes to e7 it kicks the white knight to the not very great b3-square. Black could of course play 7 . . . d6 as the rush with the g-pawn does not fit with Ad3, but then the queen might not


belong at c7. (In the lines where White plays i.d3 Black does not play Wi c7 once in the lines given by Emms in his excellent book Play the Najdorf Scheveningen Style. More on d6 set-ups in Game 33.) . 8.�b3 i.e7 9.f4

Threatening e4-e5 , which will kick the knight to a bad square and gain the e4-square for the white pieces. 9 • • • d6 10.a4!

A very interesting idea. White wants to play a4-a5, which will seriously cramp Black's queenside. IO • • • b6

Almost forced. If Black allows a4-a5 White is clearly better.

1 0 . . . ttJ c6 l l .a5 is just clearly better for White. Emms' idea 1 1 . . . ttJ d7 planning to play ttJ c5 and maybe win the a-pawn, does not work at all because of 1 2.i.e3 ttJc5 13 . ttJa4! Refuting the idea. 1 3 . . . ttJxd3 14.cxd3 And White is controlling the cfile and b6, and is much better.

1 l .e5! Without this move I would not

recommend this line for White. White has a lead in development and is trying to blow Black away immediately. After normal moves like i.d2, Wie2, Whl etc. Black would be happy, as he has reached his desired set-up and kicked White's bishop to the b3-square. Also, without I I .e4-e5 it is doubtful White should play 1 0.a4 at all. 1 l • • • �fd7

Almost forced. After the dangerous 1 1 . . .dxe5 1 2.fxe5 the pawn is poisoned ( 12 . . . Wixe5? 1 3 .i.f4 Wih5 14 .i.e2 Wih4 ( 14 . . . Wig6 1 5 .i.f3) 1 5 .i.g3 Wig5 1 6.i.f3 wins for White.) and Black must play 1 2 . . . ttJ fd7. I recently made an attempt to revive this line for Black, but (fortunately for us) I could not find a satisfactory reply to 13 .'lWf3!? Emms dismissed this line in his great book Sicilian Kan because of 1 3 . . . ttJxe5 ( 1 3 . . . 0-0 is too dangerous after 14 .Wih3 g6 ( 14 . . . h6 1 5 .i.xh6 is winning) 1 5 .i.h6 :9:e8 16 .i.g7! A brilliant attacking idea my little slave Fritz found while I was cooking. Now White is winning.) 14 . Wixa8 i.b7 1 5 .i.b5t axb5 1 6.'lWa7

With the verdict unclear. I, and maybe more notably Fritz, do not agree with this evaluation at all. In the modern age where computer programs are very strong (And everyone has one!) such positions can be worked out to a definite conclusion. I have not found a line where Black gets


sufficient compensation. Some sample lines:

1 6 . . . 0-0 17.axb5 (Taking the c6-square away from the queen and knight. 17 . lt'lxb5? j,c5t wins for Black. 1 7.i.f4 i.c5t 1 8 .lt'lxc5 %Vxc5t 19 . 'it>hl %Vc6 20 J'U2 is possible, but risky.) 17 . . . It'l bc6 1 8 . lt'l d5! exd5 1 9.bxc6 �xc6 ( 1 9 . . . lt'lg4 20J �f4 �xc6 2 1 .�a4.) 20. lt'ld4 i.c5 2 1 . i.e3 �c7 22.c3 It'lc4 (22 . . . lt'l g4 23.i.f4.) 23.i.f4 and White is better. This, of course, awaits future practical tests. If White does not like this he can just play 13 .�g4 g6 14 .j,f4 lt'lc6 1 5 .i.e4 i.b7 1 6J !ae 1 with a very active position. Please notice that the e-pawn is immune: 1 6 . . . lt'l cxe5 (Or 1 6 . . . lt'l dxe5 17 .i.xc6t) 17 .i.xb7! when White is winning material. 12.exd6 .hd6

1 2 . . . %Vxd6 transposes afrer 1 3 . lt'l e4 �c7. 13.�e4 Ae7

1 3 . . . i.b7 14 . lt'lxd6t �xd6 1 5 .j,e3 and White has the bishop pair in an open position. Black is solid but White's advantage is quite clear. 1 5 . . . lt'l c6 16.�g4 0-0-0 17 J!fdl It'lf6 1 8 .�e2 It'l b4 and now instead of 19 .i.c4 which led to a draw in Trabert - Vasilev, Kavala 200 1 , 1 9.a5 b 5 20.i.xb5 is almost decisive. 14.f5!?

White should not rest for a second - he must use his better development before Black consolidates. 14 . . . exf5

14 . . . e5 1 5 . %V g4 is terrible for Black. Black never had a chance in Sulskis -Emms, Gausdal 1 995 : 1 5 . . . 'it>f8 16 .i.g5 f6 1 7.i.e3 It'lc6 1 8 J :!f3 It'lb4 19 J �g3 E:g8 20.i.h6! It'lxd3 2 1 .i.xg7t 'it>e8 22.cxd3 �c2 23.�h5t 'it>d8 24.�f7 E:e8 25 .lt'lxf6 1-0. 15.Ji4!

White is very active and keeps developing with tempo. 1 5 . E:xf5 It'l e5 does not promise anything. 15 ... �e5

1 5 . . . �c6 1 6. lt'l d4 and 1 5 . . . �d8 1 6. lt'l d6t is very scary for Black. 16.�c3?!

1 6. lt'l g3 is much more aggressive and according to my analysis the complications favour White. White has to be prepared to sacrifice a piece in several lines here. Black has a choice between the solid 1 6 . . . 0-0 afrer which 17 .lt'lxf5 It'l bc6 1 8 .lt'lxe7t �xe7 1 9.i.e4 Ieads to a position where White has the bishop pair


and the initiative, or the 'greedy' 1 6 . . . g6!, which demands very aggressive play from

White. He has two lines to choose from:

a) 17 Ji:e 1 I think this is best. 17 . . . .!tJbc6 ( 1 7 . . . .!tJ bd7 is not good, and after 1 8 . .!tJxf5! gxf5 19 .i.xf5 f6 20.�h5t @d8 2 1 . �adl White has a big attack.) 1 8 . .!tJ d4!

Improving the position of the worst placed piece. Black now has three tries:

al ) 1 8 . . . i.c5 1 9.c3 .!tJxd4 20.cxd4 !xd4t 2 1 . @hl is good for White.

a2) 1 8 . . . f6 1 9 . .!tJ dxf5 !xf5 ( l 9 . . . gxf5 20. �h5t @d8 2 1 . �adl !c5t 22. @hl �g7 23.!h6 �a7 24.i.xf5t @c7 25.!g7 !xf5 26 . .!tJxf5 looks good for White) 20 . .!tJxf5 gxf5 2 1 . �h5t @d8 22.!xf5 i.c5t 23 . @hl �e7 24. �ad lt !d6 25 . �xd6t �xd6 26. �dl i s much better for White.

a3) 1 8 . . . i.b7 1 9.�e2 f6 20 . .!tJ e6! and White seems to hold an advantage.

b) 17.i.xf5 gxf5 1 8.�d5 f6 1 9.i.xe5 fxe5 20 . .!tJxf5 ixf5 2 1 . �xf5 is given as good for White by Emms but after 2 1 . . . �a7! this is not clear at all. 16 ... 0-0 17.�d5 .ie5t IS.�hl Wb7 19..ie2 �bc6 20.Wel .te6 21 .e4 gadS

Black has no problems after 2 1 . . . .!tJ g6! 22 . .!tJxc5 bxc5 23.!d6 �fd8. 22.l::Ml �hS 23.Wc3 f6 24.�xc5 bxe5 25 . .ie3 Wa7 26J�d2 �UeS 27.gfdl :gd7 2S.h3 f4 29 . .ixf4 �d4 30..ih5 gbS� 31 .b4 hd5 32.cxd5 gxd5 33.bxe5 �ec6 34..ixbS

Game over. 34 •.. WxbS 35 . .if3 �xf3 36.Wxf3 1-0

I have given up this variation as Black because of this line.

Game 29 van der Wiel - Nijboer Holland 1 996

1 .e4 e5 2.�f3 e6 3.d4 cxd4 4.�xd4 a6 5.�c3 We7 6..id3 .ie5!�

This is a very typical thing for the Kan - after Black encounters problems in one line he changes the move order and tries to avoid the problem. 7.�b3 .ie7

If White now continues like in the Acs' game Black can just keep the knight at g8 for a move or two and thereby sidestep the e4-e5 advance. S.Wg4!

White immediately attacks g7. This is the downside to postponing iLl f6. 8 . . . g6

Black has two alternatives: a) s . . . iLlf6 9.�xg7 l'%gS 1 O .�h6 l'%xg2

1 1 . .if4 and the bishop comes to g3, which is horrible for Black.

b) s . . . if6 A clumsy move. White just develops naturally and has a better position. 9 .id2 iLlc6 1 0.0-O d6 1 1 .'lWg3 and White is slightly better. 9 • .ig5!?

White uses the fact that the Black queen is at c7 to try to exchange the black squared bishop. As all (!) Black's pawns are currently located on light squares, Black will have weak dark squares for a long time to come. Please note that this is not possible in the fashionable line: l .e4 c5 2. iLl f3 e6 3.d4 cxd4 4 .iLlxd4 a6 5 .id3 ic5 6. iLlb3 ie7 7.�g4 g6 and there is no ig5 due to the queen at dS. 9 . . . h5

9 . . . d6 is more sensible. Hector has tried two moves in this position:

a) 1 0.0-0-O!? Hector recently switched to this move. It is probably best, and White does seem to have good chances of getting an advantage. Hector - Agrest, Aarhus 2003 continued 1 0 . . . iLl f6 1 1 . 'IW f4 iLlbd7 12 . l'%hel ( 1 2 .<;t>bl also looks promising e.g. 12 . . . b5 ( 1 2 . . . e5 1 3 .�d2 b5 14.a3) and now 1 3 .a3 as it is not certain the rook belongs on e 1 .) 1 2 . . . e5 13 .'lWd2 b5 ( 1 3 . . . iLlb6?! 14 .ie2 Wins a pawn or forces the knight to go back again.) 14 .ixb5 (Very typical of Hector, but I will recommend 14.a3(!) instead. After 1 4 . . . ib7 1 5 . <;t>b l 0-0-0 1 6.f3 White is better - he enjoys more space and has a safe king.) The game concluded 14 . . . axb5 1 5 . iLlxb5 �bS 1 6. iLl xd6t ixd6 17.�xd6 �xd6 I S . l'%xd6 l'%a6 19 . 1'%xa6 ixa6 20.ixf6 iLlxf6 2 1 . iLlc5 ics 22.c4 <;t>e7 23. <;t>c2 iLld7 24. iLl d3 ia6 25 .<;t>c3 l'%cS 26.b3 <;t>e6 27.a4 iLlc5 2s. iLlxc5t !!xc5 29. !!dl !!cS 30.f3 f5 3 1 .a5 fxe4 32.fxe4 ib7 33. !!el ia6 34. !!dl ib7 35 . l'%el ia6 36. l'%e3 h5 37.h4 l'%c5 3SJ:1g3 <;t>f6 39. l'%d3 !!xa5 40. !!d6t <;t>e7 4 1 . !!xg6 ics 42. !!g5 ig4 43. <;t>b4 !!a2 44. l'%xe5t <;t>d6 45 .!!d5t <;t>c6 46.!!c5t <;t>b6 47. l'%b5t <;t>c6 4S. l'%c5t <;t>b6 49.!!b5t <;t>c6 lh-Y2.

b) 10 .ixe7 �xe7 (An earlier Hector - Agrest game went 1 0 . . . iLl xe7 1 1 .0-0-0 e5 12 .�g5 ie6 13 .ie2 iLl bc6 14 .�d2 and here Ziegler in ChessBase Magazine stated that White is slightly better.)

Experts vs. the Sicilian I S7

1 1 .0-0-0 tL'l f6 12 .�g3 tL'lbd7 13 .ie2 eS 14 . �e3 0-0 I S .g4 bS 1 6.gS tL'leB 17. tL'l dS 'lWdB I B .h4 ib7 1 9.hS ixdS 20JhdS tL'l b6 2 1 . Eid3 tL'l c4 and the game was later drawn in Hector - Mortensen, Copenhagen 1 996. 10.Wh4 d6 1 1 .0--0-0 �c6

White also has good attacking chances after 1 1 . . .f6 12 .id2 gS 13 . � g3 h4 14.�e3. 12.ixe7 Wxe7 13.Wg3 h4 14.We3 g5 15..te2 .id7?!

A mistake but Black's position is full of holes anyway. The text allowed . . . 16.Wb6!

More or less winning a pawn by force. The b-pawn is attacked and White is planning to double rooks on the d-file. 16 ... �f6?

The lesser evil was 1 6 . . . �dB 17 . 'lWxdBt ( 17. tL'l a4!? But not 17.'lWxb7?? Eia7!) 17 . . . EixdB I B . Eixd6 and White is a clear pawn up.

17.Wxb7 Wd8 18.�d4 �a5 19.Wb4 We7 20.�b3 �b7 21.�d2! gb8 22.�e4 The rest is easy. 22 ... h3 23.g3 J.c6 24.�xd6t �xd6 25.Wxd6 Wxd6 26.gxd6 he4 27.�xe4 �xe4 28.gd4 f5 29.a �e5 30.ge4 �d7 31 .gdl �f6 32.ge6 i>f7 33.gdd6 gbe8 34.J.e4 £4 35.gxf4 g4 36.fxg4 �xg4 37.gxe6 �xh2 38.gxe8t 1-0

Notice how weak Black's dark squares became after the exchange of the black squared bishops.

Game 30 Predojevie - Fogarasi Budapest 2004

1 .e4 e5 2.�a e6 3.�c3 A smart way to sidestep different

lines like the tricky 4 . . . �b6 and the Pin-variation. Black cannot exploit this move order in any way as both the White systems recommended in this chapter involve playing the knight to c3 . 3 ... a6 4.d4 cxd4 5.�xd4 We7 6.i.d3

�f6 7.0-0 i.e5 8.�b3 i.a7!? Another way for Black to retreat the

bishop. Black is playing in similar fashion to 1 .e4 cS 2. tL'l f3 e6 3.d4 cxd4 4. tL'lxd4 a6 s .id3 ics 6. tL'l b3 ia7. It is White's job to make sure that he gets a worse version! 9.i>hl !

If White chooses to play 'IW e2 and ie3 then Black simply gains a good line


from the 5 .i.d3 i.c5 6. lLlb3 i.a7 system. Once again this shows that it is necessary to have a broad knowledge of the various lines in the Sicilian. 9 ... d6

9 . . . h5 1 0.f4! lLlg4 1 1 .'lWf3 b5 1 2.e5 i.b7 1 3 . lLl e4 shows how ineffective the bishop can be at a7.

9 . . . lLlc6 1 O.f4 d6 just transposes to the game. 9 . . . b5 might be premature after 1 O.a3 lLlc6 1 l .f4 d6? 12 .i.xb5! 0-0 13 .i.e2 Black was just a pawn down and White went on to win in Semenova - Drmic, Harkany 1 997. 10.£4

1 0.i.g5 is also possible. If White does not like the text he can investigate this idea.

1 o .. .ti:Jbd7 Black can also play 1 0 . . . lLlc6. White

now plays 1 1 . Wf e2! and seems to be a bit better as Black has problems in finding a safe haven for the king. 1 1 . . .i.d7 (Not

the optimal square for the bishop but the more active and thematic 1 1 . . .b5 does not solve Black's problems after the standard plan 12 .i.d2 0-0 1 3 .a3 i.b7 14 . E:ae l lLl e7 1 5 .e5 . This is White's normal play/plan against the Kan, and if you do not know what to do, this is often the recommended course! lLl fd5 1 6.exd6 lLlxc3 1 7.i.xc3 Wfxd6 1 8 .i.b4 'lWd7 19 . E:dl i.d5 20.i.xe7 'lWxe7 2 1 . i.xh7t 'it>xh7 22. E:xd5 and White won in Wiese - Guimaraes, Lisbon 200 1 .) 1 2.i.e3! Only now, when Black has messed up his set-up with i.d7, does White agree to this exchange. The game Ecsedi - Feher, Hungary 1 995 concluded quickly: 1 2 . . . i.xe3 13 .Wfxe3 b5 14 . E:ae l (!) O-O-O? 1 5 .a4 lLla7 16 .E:al g5 1 7.axb5 lLl g4 1 8 .Wfgl gxf4 1 9 .b6 'lWc6 20.bxa7 lLl e3 2 1 . E:xa6 'lWa8 22. E:xf4 1-0. Admittedly not a great game, but White is better!

Some Black players (maniacs) will go 1 0 . . . h5!? This is not very trustworthy and if White just sticks to the scheme ('IW e2, i.d2 E:ae l and e5) he will get a good position with nice attacking chances. 1 1 .J.d2

White is making room for the rook at e 1 . After e4-e5 the bishop will be good on the cl-h6 diagonal if Black takes on e5. And if Black does not take, and instead retreats the knight, this bishop can often find a nice square at c3 (after lLlxd5 and exd6) . In both cases the bishop will be helpful for White's attack.


1 1 • • • 0-0 12.Y!Ye2 ge8?! Black does not have time for this, but

even after the better 12 . . . b5 13 J �ae 1 .ib7 14.e5 dxe5 ( 14 . . . tLld5 1 5 .exd6 �xd6 16. tLl e4 �c7 1 7. tLl a5 .ic8 1 8 .c4 bxc4 1 9 .tLlxc4 and White has the initiative.) 1 5 .fxe5 tLl d5 1 6. tLl xd5 .ixd5 17 . .ig5! White has a strong attack. By the way, this is a good position to use to "play-out" against a computer program. 13Jfae1

White is ready for e4-e5 with a big attack. 13 ... e5?!

13 . . . ib8 (preventing e4-e5) is probably better, even though it is awfully passive. White should go 14.g4 when Black is pushed back. 14.g4!

Here we go! In general White does not push his g-pawn (and often not even his f-pawn) in the lines recommended in this chapter but, when the conditions dictate

it, White must include the kingside pawns in the battle or else he will risk losing the initiative.

With g5 coming White is getting seriously involved in the fight for control of d5. 14 ... tLlm 15.g5 tLl6d7 16.tLld5 Y!Yd8 17.£5

Black is busted. 17 ... b5 18.tLlaS tLlb8 19.c4

19 . �h5 is also good. 19 . . .J.b6 20.tLlxb6 Y!Yxb6 21 .b4 J.b7 22.cxb5

22. � e3 is simpler. 22 ... d5 23.J.e3 Y!Yc7 24.gc1 Y!Yd7 25.a4 g6 26.b6 tLlc6 27.tLlxc6 J.xc6 28.b5 dxe4 29.bxc6 exd3 30.Y!Yg2 1-0

The lines with .ia7 promise White good chances if he plays precisely.

Now we move on to the current mainline in the Kan .

Game 31 Adams - Kasimdzhanov FIDE World Championship, Tripoli 2004

1 .e4 c5 2.tLla e6 3.d4 cxd4 4.tLlxd4 a6 5.tLlc3 b5!?

Black immediately starts the attack with the wing pawn, which is so typical of the Taimanov and Kan . The advance of the b-pawn creates the option of b5-b4, which increases Black's influence in the


centre. The downside for Black is that the a6-b5 pawn-phalanx can easily become a weakness after either a well timed c2-c4 or especially a2-a4. Black is only making pawn moves so White is ahead in development and, even though Black is solid, White has good chances of creating an attack. This line has been very popular in the new millennium. 6.id3 tyb6

This has become very popular and is currently regarded as Black's best chance in the Kan after 5 .lt:l c3 . As usual Black wants a normal set-up with the queen at c7, but first he tries to kick White's knight from d4 to the passive square b3. This is known in some sources as the Enhanced Kveinys variation. I will recommend: 7.ti:lf3!?

It looks almost like a beginner's move - blocking the f-pawn and what on earth is the knight doing here? Of course the knight was under attack and had to

retreat (I have not found any trace of an advantage after 7.ii.e3!? ii.c5!) . At b3 the knight is just passive and sometimes even gets in the way, but at f3 it is eyeing g5 and e5 . Black is using a lot of time to kick the knight so at the moment his development is in a sorry state, and this might give White the chance for a rapid attack with the pieces. When I was preparing this line for Black, this was the variation I was afraid of. If White can attack with the pieces instead of pawns he is also not taking the usual risks: If the attack fails the position will be approximately equal, whereas if this happens in the normal Sicilian lines, White is usually left with a lot of pawn weaknesses. 7 ... tyc7

Black has many alternatives, but I believe this is best. Black avoids allowing White to go a4-a5 with tempo. The other lines are a jungle:

a) 7 . . . lt:l c6 8 .0-0


and now: a 1 ) 8 . . . .ib7 was good for White after

9.'�e2 (or 9Ji:e l !) lLl ge7 1O .!e3 �a5 1 1 .a3 lLlg6 1 2. lLl d5 exd5 13 . exd5 0-0-0 14.dxc6 dxc6 and later 1-0 in Krakops - Bellin, Gausdal 2000.

a2) 8 .. . lLlge7 9Ji:e l ! is nasty for Black. 9 . . . lLl g6 (9 . . . 1&b8 transposes to line c below) 1 0. lLl d5!? ( 1 0.a4 allows 1 O . . . !c5! when Black is more or less OK) � d8 1 1 .a4 l:i:b8 1 2.axb5 axb5 1 3 . lLl c3 b4 14 . lLlb5 .ic5 1 5 .!e3 !xe3 16. l:i:xe3 0-0 17. lLl d6 with a clear advantage for White, Xu Yuhua - Khurtsidze, Hyderabad 2002.

a3) 8 . . . 1&b8 (planning a set-up with lLlge7-g6 and !d6) I think White's best move is 9 . l:i:e l ! (9.e5 is also very popular) . Black now has a choice between a3 1 9 . . . !d6, a3) 9 . . . lLl ge7 and a33 9 . . . d6:

a3 1 ) 9 . . . .td6 1 0 .a4! . Freeing c4 for the knight. 1 O . . . b4 l 1 . lLl b l lLl ge7 12. lLl bd2 lLl g6 1 3 . lLl c4 !c7 14 .!f1 lLlge5 1 5 . lLl cxe5 lLlxe5 1 6. lLl d4 !b7 17.1&h5 0-0 and now, instead of 1 8 . l:i:e3, which quickly led to a draw in Smirnov - Fominyh, Chennai 2004, White can keep an edge with 1 8 . .td2! attacking b4 and preparing c2-c3 .

a32) 9 . . . lLl ge7 10 .!e3! eyeing b6. 1 0 . . . d6 ( 1 0 . . . lLlg6 1 1 .a4! b4 12 . lLld5 i s very good for White.) 1 1 .1&d2 lLlg6 12 . lLld4 lLlxd4 13 .!xd4 lLl e5 14 .f4 lLlc6 1 5 .!f2 .ie7 1 6. lLl d5 exd5 1 7.exd5 This position was reached in two games:

Myo Naing - Zaw Win Lay, Bangkok 2004 was fun: 1 7 . . . lLl a7 1 8 . l:i:e3 !d7

19 . 1:i:ae 1 lLlc8 20 . .ih4 f6 2 1 . 1&e2 1&b6 22. @h1 �d8 23.!xf6! gxf6 24.1&h5t @f8 25 .1&h6t @f7 26 . .txh7 !g4 27.�g6t @f8 28.�xg4 l:i:xh7 29. l:i:g3 @e8 30.1&g8t @d7 3 1 .�xh7 1-0.

In Grischuk - Smirin, New Delhi (2) 2000 Black instead went 1 7 . . . lLl e5 1 8 .fxe5 dxe5 19 .1&e2 0-0 20 . .id4 exd4 2 1 .1&xe7 g6 22.�f6 !b7 23.d6 !d5 24.1&xd4? (White is totally winning after 24 . .ie4!) . The game concluded 24 . . . !e6 25 .l:i:xe6 fxe6 26.c4 bxc4 27 . .ixc4 h5 28 .l:i:c1 l:i:a7 29.!xe6t @h7 30.h4 l:i:b7 3 1 .d7 l:i:xb2 32.!h3 Y2-Y2.

a33) 9 . . . d6 should by answered by 1 0 . .id2! planning the typical a4 and c3 .

a4) 8 . . . lLl f6 is very provocative after 9.e5 b4 10. lLl e4 lLlg4 1 1 .!f4 �a5 Black defended successfully in Bauer - Ciuksyte, Zurich 2002 after 12 .l:i:e 1 lLlgxe5 1 3 .lLlxe5 lLlxe5 14 .1&h5 lLlc6 1 5 . lLl g5 g6 1 6.1&f3 .ie7 17.h4 0-0. But White is almost winning after 12. lLl eg5! For example 12 . . . h6 1 3 .lLlxf7! @xf7 14.h3 or 12 . . . .tb7 13 . .te4 with tremendous pressure.

a5) 8 . . . d6 9. �e2 when White has done well in practice. Spraggett - Boudy, Montreal 1 996 proceeded: 9 . . . lLl f6 10.a4! b4 l 1 . lLl b 1 Another route for the knight. 1 1 . . ..ie7 1 2. lLl bd2 coming to c4. 1 2 . . . 0-0 ( 1 2 . . . �b8 did not help Black in Baklan - Stiri, Corinth's 2004. The game concluded 1 3 . lLl c4 lLlg4 14 .!f4 e5 1 5 .!d2 5? Asking much too much of the black position 1 6. lLl e3 lLlxe3 17.fxe3 0-0


1 B .exf5 dS 1 9.e4! .tcst 20. mh1 .txfS? 2 1 .exfS e4 22 . .txe4 dxe4 23JNc4t mhB 24.'1WxcS exf3 2S.llNxc6 l-O.} 1 3 . tLl c4 Wic7 14.aS tLld7 l S . .tf4 and White is better ( 1-0, 43) .

b} 7 . . . d6

b 1 } B .O-O and now only B . . . tLl d7 has independent significance (B . . . Wi c7 transposes to 2) under Black's alternatives at move B and B . . . tLlc6 transpose to as} above} 9 .a4! is as usual the standard response. U. Ecker - E. Anka, Bayern 1 999, took an interesting course 9 . . . b4 1 0.aS IlNc7 1 1 . tLl a4!? IlNxaS 1 2 . .te3 IlNc7 1 3 .c3 Very dynamic. 1 3 . . . bxc3 1 4. 1'k 1 tLlgf6 l S . 1'hc3 WibB 1 6.Wic2 .tb7 17. tLl b6 tLlxb6 1B . .txb6 .te7 19 .1lNa4t mfB and here White has a fantastic position after 20. l':1fc l , but instead he played 20 . .tc7 and later went astray and even lost.

b2} B.a4!? b4 (B . . . bxa4 9 .tLlxa4 is better for White.) 9.aS! And Black has a choice.

9 . . . Wib7 (After 9 . . . Wic7 1 O. tLl a4 tLld7 1 1 . .te3 .tb 7 12 . tLl b6 White already has a clear advantage. The a-pawn is weak and the tLl b6 is very annoying.) . 1 0. tLl a4 tLld7 1 1 .0-0 tLlgf6 1 2.Wie2 with a pleasant position.

c} 7 . . . .tcS B .O-O .tb7 (B . . . tLl f6? 9 .eS tLl g4 1 0. tLl e4 and Black is busted.) 9 .a4 b4 1 O.aS Wic7 1 1 . tLl a4 WixaS This cannot be right. Black will miss his black squared bishop. 12 . tLlxcS IlNxcS 1 3 . .te3 Wic7 14 .llNd2 tLlf6 ( 14 . . . aS l S . .td4 f6 1 6.Wie3 and White has a raging initiative.) I S . Wixb4 tLl c6 and now instead of 1 6.llNb6 as in Meshcheriakova - Kucherova, Essentuki 2003, White has 1 6.Wic3, which gives a huge advantage. Black is missing her bishop.

d} 7 . . . ib7?! B .a4! llNaS (A sign that everything is not right with Black's setup, but B . . • b4 9.aS IlNc7 1 0. tLl a4 is much better for White.) 9.0-0 b4 1 0. tLl b l dS l 1 .eS tLld7 1 2. tLl bd2 tLlcS 13 . tLld4 IlNc7 14 .l':1el tLle7 l S . WihS g6 16 .Wih4 .tg7 17. tLl 2b3 h6 1 B .tLlxcS IlNxcS 1 9 .tLlf3 and White is better. The game concluded. 1 9 . . . tLl c6 20 . .te3 llNe7 2 1 . Wig3 md7 22.aS gS 23.c3 g4 24.llNxg4 tLlxeS 2S . tLl xeSt ixeS 26.WihS f6 27 . .td4 .txd4 2B.cxd4 l':1agB 29.f4 Wig7 30.g3 f5 3 1 . l':1eS Wig6 32.Wie2 l':1aB 33. l':1e 1 l':1heB 34.llNd2 .tc6 3S .Wixb4 l':1abB 36.Wic3 .tbS 37 . .txfS! 1-0. Ghinda - Petre, Sovata 1 999. 8.0-0


8 . • • .ib7 Again Black has alternatives: 1 ) B . . . tLl c6 is not very logical (it fits

better with �bB) 9J:!:el is a good reply. l a) 9 . . . d6 1 0.a4! The normal reply

to d7-d6. 1 0 . . . b4 l 1 . tLl a2 tLl f6 12 .i.d2 a5 (White is better after both 12 . . . d5 1 3 .exd5 tLlxd5 14.ie4 and 12 . . . �bB 13 .c3 bxc3 14 .i.xc3! i.e7 1 5 .b4 0-0 16.b5) 13 .c3 bxc3 14 . tLlxc3 and White is clearly better.

Ib) 9 . . . id6? 1 0.ixb5! l c) 9 . . . ttJf6 1 0.e5 tLlg4 l 1 .if4 f6

12.ig3! fxe5 1 3 .tLlg5 tLlf6 14.ixh7 d6 1 5 .ig6t rJ1e7 1 6.tLlf7 l:!gB 17.ttJe4 d5 I B .ttJeg5 1-0. Nikolenko - Ivanov, Moscow 2000.

2) B .. . d6 9.a4! with the usual play on the c-file after b5-b4, ttJa2 and c2-c3 .

3) B . . . tLl f6 9.e5 b4 1 0. tLl e4 and White is better. 9.ge1

9.id2!? is an interesting alternative.

White had an easy time after 9 . . . i.e7 1 0.a4 b4 l 1 . tLl a2 tLlc6 1 2.'W'el a5 1 3 .c3 bxc3 14 . tLlxc3 tLl f6 1 5 .e5 tLlg4 1 6. tLl b5 'W'bB 17.ic3! . Now all White's pieces are participating in the attack. 17 . . . 5 I B .h3 tLlh6 19 . gdl 0-0 20.i.c4 gdB? 2 1 . gxd7! gxd7 22.i.xe6t �hB 23.i.xd7 tLlb4 24.�e2 tLld5 25.id2 tLlf7 26.ixf5 ic5 27.e6 tLld6 2B.i.xh7 �xh7 29. tLlg5t �gB 30. tLlxd6 'W'xd6 3 1 .�h5 tLlf6 32. 'W'f7t �hB 33. �xb7 1-0 in Pelletier - Lobron, playchess.com 2004. 9 . . .J.c5

a) 9 . . . d6 1 0 .a4! b) 9 . . . i.e7 Epishin's most recent idea

and therefore probably the critical line. (Vladimir Epishin is the main connoisseur of this line in the Kan) . I like 1 0.id2!? with the usual plan of generating play on the queenside. 1 O . . . h4 ( l 0 . . . d6?! 1 1 .a4 bxa4 12 . gxa4 with the usual slight advantage to White even though Black managed to draw in Tseshkovsky - Epishin, St Petersburg 2004) l 1 . tLl e2 tLlf6 ( l 1 . . .tLl c6 1 2.c3 tLlf6 13 . tLlg3 bxc3 1 4.ixc3 0-0 1 5 . gel and White has a slight advantage, Tyomkin - Epishin, Kapuskasing 2004) and now instead of 12. tLl ed4 which was not very successful (even though White is slightly better) in Arizmendi Martinez - Smirin, Istanbul 2003, I think 1 2. tLl g3! , planning i.d2 and c3, gives White good chances.

c) 9 . . . tLl f6 1 0.e5! b4 1 1 . tLla4 and White will play the bishop to d2 and then go c2-c3 with an initiative. This idea should be


familiar to the reader by now! 10.tfd2

Very artistic. White can also play 10 .�d2, planning the usual a4 and a later c2-c3 opening the c-file. White has also enjoyed some success with 1 O .�g5! ? followed by manoeuvring the bishop to g3 . 10 •• .i.e7

White was threatening 1 1 . Wf g5 . 1 l .b3!? �f6 12.i.b2 d6 13.a4 b4 14.�a2 �c6

15.�d4 1 5 .c3! bxc3 16.!xc3 with the rook

coming to c 1 . White has the initiative. 15 .•. �xd4

1 5 . . . a5! is better. Black won in Zhang Zhong - Ye Jiangchuan, Taiyuan. After 16.�b5 Wfd7 17.E:adl 0-0 18 .e5? (A silly pawn sacrifice.) 1 8 . . . lLlxe5 19 .�f1 Wfc6 20.f4 lLlg6 2 1 .f5 exf5 22. lLld4 Wfc5 and Black is in control. 16.i.xd4 a5 17J��acl 0-0 18.c3 bxc3

19.�xc3 YlYd8 2o.lLlb5 e5 21 .E:c7! White is better

21. .. exd4 22.E:xb7 d5 23.YlYf4 .ib4 24.E:dl dxe4 25.ixe4 �xe4 26.YlYxe4 YlYf6 27.�c7 gad8 28.�d5YlYd6 29.tfxd4 YlYe6 30.g3 E:d6 31 .gxb4 axb4 32.�e7t YlYxe7 33.YlYxd6 YlYe2 34.gd4 h5 35.tfxb4 YlYa 36.h4 gc8 37.YlYd2 gc3 38.E:f4 YlYc6 39.�h2 E:xb3 4O.tfdSt �h7 41 .YlYdl gb7 42.YlYxh5t �g8 43.E:d4 YlYf6 44.E:d2 ga7 45.a5 g6 46.Y!Yb5 �h7 47.Y!Yb6! 1-0

It seems to me that Black has plenty of problems in this line at the moment. I doubt it will keep its current popularity.

Game 32 Svidler - Milov Biel 2000

1 .e4 c5 2.�a e6 3.d4 cxd4 4.�xd4 a6 5.�c3 b5 6 . .id3 .ib7

Another way to play. 7.0-0 tfb6

Black is trying to reach the 6 . . . Wfb6 line by a different move order. We will look at the alternatives in Game 33. 8.i.e3!?

White can also play 8 . lLl f3 probably transposing to Game 3 1 . 8 •• .i.c5 9.�ce2

Not 9 .�e2? lLlc6! . 9 ••• �f6

a} 9 . . . Wf c7 is currently regarded as Black's best chance - I think this will change after these notes are published!


White goes 1 O .a4! al ) 1O . . . b4 ( l 0 . . . bxa4 1 l .Ei:xa4 is a bit

better for White.) and here in El Arousy - Annageldyev, Dubai 200 1 , White continued I I .c3 and the game was later drawn. Instead 1 l .Ei:cl ! is very strong.

A move I found while trying to make this line work for Black. c2-c3 is coming and this will give Black big problems on the c-file. l l . . . lLl f6 ( l l . . . lLl c6 12 . lLlxe6) 12 .c3 b3!? ( l 2 . . . lLlg4 13 .!f4.) 1 3 .lLlg3 and b3 is dropping.

a2) 10 . . . lLl f6 l l .axb5 lLlxe4 12 .bxa6 lLlxa6 1 3 .lLlb5 and White is a clear pawn up.

b) 9 . . . lLl c6 is better for White after 1 O.c3 lLl f6 l l .b4 !xd4 12 .lLlxd4 Wic7 1 3 .lLlxc6 !xc6 14 .f3 0-0 1 5 .Wie2. 10.b4!

Already White is better. 10 ••. .hd4?!

Black could try 1 O . . . !xb4!? but White has a nice initiative after l l . lLl f5! .

1 l • .bd4 V!ic7 12.c4!

As usual White is generating attacking chances on the c-file (who knows, maybe this will one day teach Black players to keep the b-pawn back) . 12 ••• e5

12 . . . bxc4 13 . Ei:c1 !xe4 14. Ei:xc4 and White is attacking. 13.1b2 lLlc6

13 . . . bxc4 14Jkl d5 1 5 .exd5 !xd5 16. !xc4 !xc4 17. Wi c2 is better for White according to Ribli. 14.ab5 axb5 15.lLlc3lLld4 16.ixb5 0-0 17.i.d3

And Svidler's great technique decided the issue. 17 •. .lUdS ISJ!e1 d5 19.exd5 .bd5 20.tlJxd5 Ei:xd5 21 .Ei:c1 Wib7 22.1c4 Ei:d7 23.a3 Ei:adS 24.V!id3 e4 25.Wie3 h6 26.h3 gd6 27.ifl gd5 2SJ!c4 tlJf5 29.V!ic1 gSd6?! 30.ixf6 Ei:xf6 31.Ei:cxe4 V!lb6 32.Ei:e5 lLlg3 33.V!ic8t Ei:dS 34.V!ixdSt 1-0


Game 33 Smirin - Markowski Plovdiv 2003

1 .e4 c5 2.�f3 e6 3.d4 cxd4 4.�xd4 a6 5.�c3 b5 6.Ad3 d6

A specialty of Svidler and Markowski. 7.0-0

7 ... �f6 a) 7 . . . ib7 i s answered by the standard

8 .a4! b4 9 . lL\a2 dS C9 . . . eS is no better 1 O .lL\f5 ! lL\c6 l 1 .idl as 12.c3 bxc3 1 3 .lL\xc3 with a dear advantage in Wiersma - H. van der Spek, Amsterdam 2002.) 1 0.eS lL\c6 1 1 . lL\xc6 ixc6 12 .idl YMb8 13J:!:e1 1l;!/b7 14. lL\cl lL\ e7 l S . lL\ b3 lL\g6 1 6.f4 id7 17.1l;!/g4 and White was better and duly won in V. Malisauskas - M. Brodsky, Riga 1 995 .

b ) 7 . . . ie7 8 .a4 if6 9 .ie3 bxa4 1 0 .lL\xa4 lL\ e7 1 1 .c4 0-0 1 2.b4 lL\ bc6 1 3 .lL\xc6 lL\xc6 14J:!:b1 :1!b8 l S .YMdl id4 YZ-Y2 Kundin - Kudischewitsch,

Tel Aviv 2002. A weird decision by White, as he is much better. 8.YlYe2

The best move in my opinion. The queen is often good on this square, where it adds to the pressure on the fl-a6 diagonal, defends e4, and supports a future e4-eS . 8 .:1!e 1 is also possible. 8 .. .J.b7

8 . . . ie7 9.a4! b4 C9 . . . bxa4 1 0. lL\ xa4 ib7 1 1 .id2 0-0 12 .b4 and White has the better prospects and went on to win in Cernousek - Bernasek, Lilie Litomysl 2003) 1 0 . lL\ a2

a) 1 0 . . . eS 1 1 . lL\ f5! . With Black's central pawns on dark squares it makes a lot of sense to force the exchange of his white squared bishop. 1 L.. ixfS 1 2.exf5 as ( 1 2 . . . YMaS 13 .idl lL\c6 14.c3 bxc3 l s .ixc3 YMb6 1 6.b4 0-0 17 .YMe3 seems to be a bit better for White.) 1 3 .ibSt lL\ bd7 14 .igS 0-0 l S .c3 bxc3 1 6. lL\xc3 And White has an obvious advantage


according to Smirin in New In Chess. I tend to agree, as White has the bishop pair and free play. The black bishop does not seem to have a very bright future and if White is careful he will always be able to 'play around it' . The game continued 1 6 . . . h6 17 . .th4 tLlc5 1 8 J'!adl tLlb3 1 9 . .txf6 .txf6 20. tLld5 tLl d4 2 1 . �d3 Eic8 22. !'kl .tg5 23.Eic3 Eixc3? (23 . . . Eic5 is still good for White but Black has better chances than after the text move, which robs his knight of the d4-square) 24.bxc3 tLl b3 25 .Eidl tLl c5 26.�e2 e4 27.'�g4 .tf6 28 .Eie 1 .te5 29. Eie3 @h7 30. Eih3 g6 3 1 .f3 exf3 32.gxf3 h5 33.f4 .tg7 34.�g2 Eig8 35 .@hl .tf6 36.fxg6t fxg6 37.f5 .th4 38 . .tc4 Eig7 39.�f3 �g5 40. tLl f4 �xf5 4 1 . Eixh4 �e7 42.id3? After a well-played game Smirin throws the win away. The easiest win is probably 42 . .tb5! ' After the text move Black managed to draw. 42 . . . tLlxd3 43.�xd3 Eie4 44. tLlg2 Eixa4 45. �dl lh-lh Smirin - Markowski, Rethymnon 2003.

b) 10 .. . �b6 Markowski's latest try. l 1 .ie3 �b7 1 2 . .td2 ( 12.c3 tLlxe4 1 3 .tLlxb4 tLl f6 1 4 . .tg5 0-0 1 5 .a5 d5 16.f4 and here White quite prematurely agreed to a draw in Bakalarz -Markowski, Antalya 2004. Not the best decision as he has a promising attacking position.) 1 2 . . . e 5 1 3 .tLlb3 tLlc6 14 . .tg5 0-0 15 . .txf6 .txf6 1 6 . .tc4 .te6 17 . Eifdl The position is very unclear but later won by White in Cyborowski - Markowski, Warsaw 2004.

c) 10 . . . a5 1 1 .c3 e5 12 . .tb5t! and White

is already clearly better. V. Faibisovich - R. Rodkin, Pardubice 200 1 . 9.a4 b4 10.tLla2 d5

1 0 . . . ie7 l 1 . tLlxb4! d5 1 2.e5 .txb4 1 3.exf6 �xf6 14.c3 ie7 1 5 . .tb5t! tLl d7 ( l 5 . . . @f8 1 6.id3 e5 1 7. tLl b3 leaves White only slightly better.) 1 6 . .txd7t @xd7 17 . .te3 Eihc8 1 8 .tLlf3 .td6 And here, instead of 19 .Eife1 (Ribli - Garcia Martinez, Cienfuegos 1 972) , White should play 1 9 . .td4! with an obvious advantage. l l .e5 tLle4 12.c3 tLld7

1 2 . . . bxc3 1 3 .tLlxc3 tLlxc3 14.bxc3 �c7 1 5 . Eib l and White is better according to Smirin. Black has no way to develop his pieces in a sensible way. And where is Black's king supposed to go? 13.9 tLlec5 14.tLlxb4 a5

15.tLlbc6! Very smart. Smirin has calculated that

the discovered check is harmless. 15 •. ..bc6


1 5 . . . \Wb6 1 6.�b5 lZlb3 17 .�e3 lZlxal 1 8 .lZlxe6. 16.�xc6 Wb6 17.Ab5 �b3t?!

17 .. J �c8 is better, but even here White is clearly better after 1 8 .�e3 E:xc6 19 .b4! E:c7 20. E:ab l \Wa7 2 1 . �hl �e7 22.bxc5 �xc5 23.�g5 0-0 24.f4 with a nice kingside attack in the making. 18.Ae3 Ac5 19.�ael gc8 20.i.xc5 Wxc5t 21 .i>hl!

A computer move. 21. . . gxc6 22.Wdl!

The rest i s easy. White i s just a pawn up. 22 ... gc7 23.Wxb3 0-0 24.Wdl �b6 25.£4 g6 26.�a i>g7 27.gh3 h5 28.gd3 �c8 29.f5

A forceful shift of gear. 29 .. . exf5 30.gxd5 Wn 31 .e6 Wxb2 32J�d7 E:xc3 33.e7!

Winning a piece. 33 ... �xe7 34.Wd4t 1-0

Game 34 Ponomariov - Gallagher Biel 2000

1 .e4 c5 2.�a e6 3.d4 cxd4 4.�xd4 a6 5.�c3 b5 6.J.d3 Ac5

Another way to try to reach the positions with White's knight stuck at b3.

The alternative is 6 . . . �b7 7.0-0 and apart from 7 . . . 'lWb6, which we looked at in Game 5, Black has the following options:

1) 7 . . . 'lWc7 (In combination with b5 and �b7 this is begging for trouble. The position often arises via 5 . . . 'lWc7 too.) 8 J l:el ! ? Getting ready for a knight jump to d5. Now we have:

l a) 8 . . . lZlc6 9 .lZlxc6 \Wxc6 (9 . . . �xc6 1 0. lZl d5! and 9 . . . dxc6 1 0.a4 is better for White) 1 O.a4 b4 l 1 . lZld5 lZl f6 12 .id2 ic5 1 3 .d and White has the initiative.

I b) 8 . . . d6 9.a4! bxa4 (9 . . . b4 1 0. lZl d5 ! A typical sacrifice even though 10 . lZl a2 with the usual plan of attacking the queenside is safe and good, 1 0 . . . exd5 l 1 .exd5t lZl e7 and here instead of 1 2. lZl fS which looked unclear in Stocek - Protaziuk, Zagan 1 995, I prefer 1 2.�g5 g6 13 .'lWe2!? with an interesting attacking position.) 1 0 Jha4 White's play in this game is very instructive: 10 . . . lZl f6 1 1 . E:c4 'lWd8 12 . E:b4 'lWc7 1 3 .�g5 �e7 14 .ic4 lZlc6 1 5 . lZlxc6 ixc6 1 6.�xf6 gxf6 17. E:b3 E:g8 1 8 .'lWh5 E:g6 19 .'lWxh7 fS 20. lZl d5 1-0 Ribli - Tokaji, Hungary 1 967.


l c) 8 . . . b4?! 9 . tLl d5! (The alternative 9 . t2ke2 led to a White success after tLlf6 10 .tLlg3 �c5 I l . tLlb3 �d6 12 .�d2 tLlc6 1 3 .c3 a5 14 . tLld4 �c5 1 5 .�e3 tLl e5 1 6. tLl b5 'lWb6 17.�xc5 �xc5 1 8 .cxb4 1-0 in Tiviakov - Anand, Tilburg 1 992 but the text move is more forceful.) 9 . . . exd5 1 0.exd5t md8 1 1 .'lWh5! and White's attack is very strong.

I d) 8 . . . �d6 rapidly led to a disaster for Black. 9 . tLl f3 b4 1 O . tLl a4 tLlc6 1 1 .�e3 tLle5 12 . tLlxe5 �xe5 13 .'lWh5 tLlf6 14.'lWh4 'lWc6? 1 5 .�g5 d6? 1 6.f4 1-0 Krakops - Kunte, Zagan 1 997.

I e) 8 . . . �c5! ? 9 .tLlb3 �e7 10 .'lWg4 g6 1 1 .�g5 is similar to Game 2 .

I f) 8 . . . �e7 and in Emms - Franklin, West Bromwich 2004, White tried the adventurous 9.e5 (9 .�e2!? is a good alternative.) and after 9 . . . b4 1 O.tLla4 tLlc6 I l .tLlxc6 hc6 12.b3 h5 1 3 .�e4 h4 14 .�f3 :1:ic8 1 5 .�d2 tLlh6 1 6.:1:iac1 tLlf5 17.c3 bxc3 1 8 .:1:ixc3 he had gained a big advantage. The game concluded: 1 8 . . . 0-0 1 9 .94 hxg3 20.hxg3 �xe4 2 1 .�xe4 'lWa7 22.mg2 d5 23.exd6 tLlxd6 24.�f3 �f6 25.:1:id3 :1:ifd8 26.�a5 1-0.

1 9) 8 . . . tLl f6 is as usual answered by 9.e5! . Black went 9 . . . tLld5 in Haba - Voloshin, Plzen 2003. After 1 O .tLlxd5 �xd5 1 1 .�g4 tLlc6 12 .tLlxc6 'lWxc6 13 .�d2 h5 White could have gained a clear advantage with 14 . 'lW g5 when Black has problems developing. In the game White played 14 . �h3?! and still won with some luck.

2) 7 . . . d6 8 .�e2! 2a) 8 . . . tLlf6 Transposes to Game 33. 2b) 8 . . . tLlc6 9 .tLlxc6 �xc6 1 O.a4! with

a clear advantage. 2c) 8 . . . tLld7 9.a4 bxa4 10 .tLlxa4 and

White is a little better. 3) 7 . . . tLle7 8 .�g5! has long been

known to be good for White. Both 8 . . . h6 and 8 . . . f6 should be answered by 9 .�e3 when Black's structure is weakened and White has good attacking chances.

4) 7 . . . tLlc6!? 8 .tLlxc6 4a) 8 . . . dxc6 9.e5 (9.a4!) tLle7 1 0.'lWh5

�c7 1 1 . :1:ie l tLlg6 12 .�xg6 fxg6 13 .'lWg4 �f7 14. tLl e4 �f5 1 5 .�h4 c5 1 6. tLl d6t �xd6 17.exd6 and White went on to win in Emms - Crouch, England 1 997.

4b) 8 . . . �xc6 9J'ie l ! 4b l ) 9 . . . tLlf6 1 0.e5 is slightly better

for White after 10 . . . b4 I l . tLl e4. Black has some problems with his queenside pawns.

4b2) 9 . . . �b8!? 1 O .a4! b4 1 1 . tLld5 �d6 12 .'lWh5 �xd5 1 3 .exd5 tLlf6 14 .'lWg5 mf8 A concession. The rest of the game is instructive - White goes very hard after Black's king. 1 5 .dxe6 dxe6 16 .�h4 me7 17.a5 :1:ic8 1 8 .�d2 b3 19 .�c3 bxc2 20.�xc2 �b5 2 1 . :1:iadl :1:ic4 22.'lWh3 tLlg4 23. :1:ixd6 mxd6 24.�d3 :1:ixc3 25.bxc3 'lWg5 26.�g3t me7 27.�e2 �d2 28. mf1 h5 29.h3 :1:ic8 30. :1:idl 1-0 Tiviakov - Perez Candelario Malaga 2003 .

4b3) 9 . . . d6 1 0.a4 is good for White. 1 0 . . . b4 is answered by I l . tLld5!


4b4) 9 . . . tUe7 1O .�h5 tUg6 1 1 .a4 b4 12 . tU d5 .td6 and now, instead of 1 3 . .!e3 which was later drawn in Ponomariov -Dao Thien Hai, New Delhi 2000, I prefer 13 .'!d2!? 7.tLlb3 Ae7

8.Wg4! g6 8 . . . tUf6 9 .�xg7 �g8 10 .�h6 tUc6

1 1 . .!f4 and Black has no compensation for the pawn. 9.We2

Back to the ideal square. Compared to the line 1 .e4 c5 2. tUf3 e6 3.d4 cxd4 4. tUxd4 a6 5 . .!d3 .!c5 6. tUb3 .te7 7.�g4 g6 8 .We2 Black has played b7-b5 too early and White can punish this with a well-timed a2-a4. 9 ••• d6 10.0-0 M?!

A strange move but White was planning to hit this pawn anyway with a2-a4 (as usual) . 1 l .tLldl

1 1 . tU b 1 is also interesting.

1 l •. .i.b7 12.a3! Another way to attack the queenside.

12 ••• tLlc6 1 2 . . . bxa3 13 . �xa3 is slighdy better for

White. 13.i.d2 bxa3

1 3 . . . a5 is answered by 14 . .!b5. 14J:gxa3

According to notes by Ribli in ChessBase Magazine, White is now clearly better - I agree. 14 ••• We8

Defending a6. 15.tLle3 f6 16.tLle4 We7 17.tLlca5

Now Pono wraps it up nicely. 17 ••• tLle5 18.tLlxb7

The bishop will be sorely missed. 18 ••• tLlxd3 19.cxd3 Wxb7 20.tLlaS Wd7

20 . . . �xb2 2 1 . �b3 �a2 22. �b7 with an attack. 2U :gb3 l1!:c8 22.tLle4 Ad8

23.J.h6! d5 24.tLlb6 .b:b6 25.gxb6 Wa4 26.Wa tLld7 28.dxe4 Wa5 29.gb7


Wb5 30.Wf4 e5 31 .Wd2 �e5 32.fia7 g5 33.Wd6 1-0

And here I conclude the survey of the Kan and move on to the Taimanov.

Part II: The Taimanov

Against the Taimanov I recommend the following system: 1 .e4 e5 2.�S e6 3.d4 cxd4 4.�xd4 �c6 5.�c3 We7 6-Ae3 a6 7 • .td3. This is one of White's most aggressive systems. In no time White finishes his development and is ready for action. Plaskett's remark in his book from 1 997 The Sicilian Taimanov is still valid: 'It is possible for great violence to occur very early on in these games as you will soon gather' . A good rule of thumb is: If you spot an attacking idea - go for it!

In Game 34 we will investigate 7 ••• b5. This move has a very good reputation. In Games 35-39 we will look at Black's different moves afrer 7 ••• �f6 8.0-0. In Game 40 we cover the lines afrer 5 ••• a6.

Game 34 Hector - Lindberg Umea 2003

1 .e4 e5 2.�S e6 3.d4 cxd4 4.�xd4 �c6 5.�c3 We7 6.!e3 a6 7 -Ad3 b5

This has a very good reputation. Once, when I was preparing to play against Hector, I noticed that the line he plays against this variation is actually very

annoying for Black. So here we go. 8.�xc6

The move b7-b5 is almost always answered by lDxc6 in the Taimanov. 8 ••• Wxc6 9.0-0 Ab7 10.�bl (!)

Preparing £2-f3! 10 ••• �f6 ll .S!?

I like this idea. White just fortifies the centre and gets ready to attack Black's queenside with a4. Just like we did against many lines in the Kan .

l l •• ..te5! I think this move is best. Black has two

alternatives. a) 1 1 . . .�c7 is risky. 1 2.a4 b4 1 3 .lDe2

i.e7 14 .Wd2 0-0 1 5 J :ife 1 i'HcS And now instead of 16 J !ad1 ? which allowed 1 6 . . . d5 ! 1 7.e5 lD e4!! and Black was better and went on to win in Hector - Cramling, Malmo 200 1 , I like 16 .lD d4! . For instance 1 6 . . . d5 17.e5 lDd7 ( 17 . . . Wxe5 l S .i.f4 Wh5 1 9J !e5 �h4 20.g3 �h3 2 1 . i.f1 and wins) And White has a very nice

French structure. Do not get confused by the chess programs' evaluations - they do not understand the position! Sample line: 1 8 .i.g5! ? i.c5 1 9 .tLlb3 i.f8 20.YNe2 tLlc5 2 1 . tLlxc5 i.xc5 22.f4 h6 23.i.h4 i.d4 24J'iab l and White has a free hand on the kingside.

b) l l . . .d6 12 .YNd2 ie7 1 3 .a4 bxa4 14 J'l:xa4 0-0 1 5 J'l:fal E:fc8 and 16 .YNe2 ( 1 6. tLl e2!? might be better. For instance, 17 . . . d5 1 7.e5 tLld7 1 8 .i.g5 i.f8 19 . E:h4 h6 20.i.f4 with an interesting attacking position. ) . 1 6 . . . d5 1 7.exd5 tLlxd5 1 8 . tLl xd5 YNxd5 19 .i.xa6 (White could also force a good ending with 19 .i.e4 YNd7 20.YNd3) Hector - Tozer, Copenhagen 2002, concluded 1 9 . . . i.xa6 20. E:xa6 E:xa6 2 1 . E:xa6 YNe5 22. E:a7 i.d6 23.f4 YNxb2 24. YNa6 YNbl t 25 .i.gl YNb8 26.g3 h6 27.c4 ic5 28. E:b7 YNd6 29.YNxd6 i.xd6 30 J'l:d7 E:c6 3 1 . <;t>g2 g5 32. <;t>S <;t>g7 33 .h4 gxf4 34.gxf4 h5 35 .i.d4t <;t>g6 36. <;t>e4 f6 37.f5t 1-0. 12.'1flYel .he3 13.Wxe3 We7

A strong move according to Ribli. Alternatives:

a) 1 3 . . . b4 14 . tLl e2 0-0 1 5 J Hcl ! We have seen this theme before! 1 5 . . . e5 1 6.c3 YN d6 17 . tLl g3 and White was better in Short - Rogers, Manila 1 992.

b) 13 . . . 0-O? 14.e5 tLld5 ? 1 5 . tLl xd5 YNxd5?? 16 .i.e4 YNxe5 1 7.i.xh7t <;t>xh7 1 8 .YNxe5 wins.

c) 1 3 . . . d6 14.a4 b4 1 5 . tLl a2 YNc7 1 6. YN d2 transposes to the main game. 14.a4

1 4.e5 tLld5 1 5 . tLl xd5 i.xd5 is equal according to Ribli - and Ribli is almost always right! 14 ... b4 15.�a2 a5 16.d! bxd 17.�xc3 0--0 18.�b5

1 8 .e5 tLld5 1 9 .tLlxd5 i.xd5 does not give anything. 18 .. . Wb8 19.e5

White could also try 1 9 . E:ac l !? with the more pleasant position.

1 9 .E:fe l E:c8 20. YNd2 E:c5 2 1 . i.f1 i.c6 22. tLl d4 and draw agreed in Lutz - Ribli, Germany 1 996, is not what we want! 19 ... �d5 20.We4 f5 21 .Wd4!

After 2 1 .exf6 tLlxf6 a draw was agreed in Z. Almasi - Leko, Groningen 1 995 . 21 . . .J.e6 22.�d6 �b4!

The only chance. If White is allowed to play ib5 he will take over the c-file with an easy win. 23JUdl �xd3 24.E:xd3

I think this position is great for White. He has all the play and can slowly prepare


a kingside assault with a transfer of the queen to the kingside followed by f4 and :B:g3. If White is careful Black will not have any counterplay. 24 .. J:�a6 25.Wc3 Wb6 26.b3 :B:a7 27.:B:cl gbS 2S.Wdl h6 29.ge4

29. :B:dc3! �h7 30.'lWe1 :B:m 3 1 .'lWh4 :B:aa8 32.f4 'IW d4 33. :B:g3 looks promising. In the game White starts to drift. 29 .•• s!?h7 30.h3 :B:m 31 .s!?h2 gaaS 32.h4 gabS 33.gdc3 J.d5 34.gf4 .bb3 35.g4 J.d5 36.gxf5 exf5 37.h5 ie6 3S.gcl Wb2 39.Wxb2 gxb2t 40.s!?g3 :B:gS 41.gd4 g5 42.hxg6t :B:xg6t 43. s!?f4 :B:gS 44J�e7?

44. gh! ! and White should not lose. 44 ... gb4 45.�b5 :B:bS 46.s!?e3 s!?g6 47.f4 s!?f7 4S.:B:a7 h5 49.gxa5 E:b3t 50.s!?f.2 h4 51 .:B:dl :B:b2t 52.s!?e3 h3 53.�d4 h2 54.�xe6 gSb3t 55.s!?d4 E:bl O--1 .

I t i s noteworthy that Ribli does not play this line anymore.

Game 36 Hector - Pogorelov Copenhagen 2004

1.e4 e5 2.�f3 e6 3.d4 cxd4 4.�xd4 �e6 5.�c3 We7 6.J.e3 a6 7.J.d3 �f6 S.o--o id6

8 . . . bS just transposes to Game 3S after 9 .tDxc6 'lWxc6 1 0.f3 J.b7 1 0. �h l . We will deal with Black's other options: 8 . . . hS, 8 . . . tDxd4, 8 . . . tD eS and 8 . . . d6 in the following games.

9.f4

A very aggressive move. White is trying to punish Black for the extravagant bishop move. 9 .. .!e5

Once again we dive into a heavy theoretical minefield.

a) 9 . . . tDxd4 1 0.eS! al) 10 . . . ,ixeS is risky. l UxeS 'lWxeS

12 .'lWd2 with a further split: a l a) 12 . . . bS 13 . :B:ae 1 tDg4 14 .,if4

'lWcS l S . �h1 ,ib7 1 6.h3 tDf6 17.,ieS and White has a huge initiative.

a lb) 1 2 . . . tDg4 13 .,if4 'lWcS 14 . �h1 fS ( 14 . . . eS l S . :B:ae 1 ) l S . tD a4 'lWc6 1 6.'lWb4 bS 1 7.'lWxd4 ,ib7 1 8 . :B:f3 bxa4 19 .'lWxg7! and White is much better. This is an improvement on 19 .,ixfS which ended in a draw in Vavra - Bunk, Bayern 1 999.

a 1 c) 12 . . . tDdS? 1 3 .tDxdS 'lWxdS 14.c4 'lWd6 l S .'lWf2 tDfS 16 .,ixfS exfS 17.,icS 1-0 Roger - Lemeaux, France 2002.

a Id) 12 ... tD c6 13 .:B:ae 1 0-0 ( 1 3 ... bS


14 . lt>hl !? .tb7 I S . .igS ! (improving on I S . .tb6 from Senff - Miezis, Oslo 2003) I S . . JMb8 16 . .ixf6 gxf6 17. ttJdS with a fantastic attack.} 1 4 . .ib6 WihS I S . gxf6! gxf6 And now instead of 1 6. ttJ e4 as in Gatcia - Lukov, La Pobla de Lillet 1 996, White can play 16 . ge3! ttJ eS 17 . .td4!! with a winning attack.

a2} 10 . . . ttJ c6! ? 1 1 .exf6 ( 1 l .exd6 is uncleat.) 1 1 . . .gxf6 12.WihS ( 12 .Wig4 .ie7 13 .Wig7 gfB 14. ttJ e4 f5 I S .ttJf6t .ixf6 16. Wixf6 is very uncleat. ) . 12 . . . .ie7 ( 1 2 . . . f5 13 . .ixfS WiaS 14 . .tg4 WixhS IS . .txhS .ib4 1 6. ttJ a4! and White is clearly better.) 1 3.f5 ttJ eS 14. gadl bS and instead of I S . .ie4 which turned out badly for White after I S . . . .ib7 16 . .id4 0-0-0 17 . .ixb7t VNxb7 1 8 .fxe6 fxe6 1 9 . .ixeS fxeS 20.WixeS ghg8 2 1 . ttJ e4 gg4 22. gd4 gdg8 23.g3 Wic7 24.Wixc7t It>xc7 in Hector - S. Salov, Copenhagen 1 997, I prefer I S .fxe6 dxe6 16 . .ie4 .ib7 17 . .td4 0-0-0 1 8 . .ixeS fxeS 1 9 . .ixb7t VNxb7 20. Wi xeS with a slight advantage.

a3} 1 O . . . .icS 1 1 .exf6 a3a} 1 1 . . . ttJxc2 and the very famous

game Azmaipatashvili -Kurajica, Strumica 1 995, continued 12.fxg7 .ixe3t 13 .lt>hl gg8 14.Wixc2 gxg7 I S . gae l .ixf4 1 6. gxf4 VNxf4 17.ttJdS Wih4 18 . ge4 gg4 1 9 .93 WigS 20. ttJ c7t It>d8 2 1 . ttJ xa8 gxe4 22 . .ixe4 WiaS 23.Wic3 Wixc3 24.bxc3 dS 2S . .ixh7 bS 26.tll b6 .tb7 27. lt>gl It>c7 28.h4 d4 29.hS dxc3 30 . .ic2 f5 3 1 .h6 .ie4 32.h7 1-0. It was rumoured that this game was preatranged, or maybe not even

played at all, but that does not change the verdict: 1 1 . . . ttJxc2 is bad.

a3b} l 1 . . .ttJf5 12 . .ixcS Wixcst 1 3 .gn is just better for White. 1 3 . . . g6 ( l 3 . . . gxf6 14 .WihS is very good for White.) 14 . .ixfS gxf5 and here I like I S . Wif3 preventing b7-bS.

a3c} l 1 . . .ttJbS I 2.fxg7! .ixe3t 1 3 .lt>hl gg8 14 . .ixbS gxg7 ( 14 . . . axbS I S .VNd3 wins for White.)

I S . .id3 .ixf4 1 6. ttJ e4 .ieS 17 .VNhS bS ( 17 . . . d6 1 8 .ttJf6t .ixf6 19 . 9xf6 and with the other rook coming to fl, White has good chances.) 1 8 . gxf7! is good for White. 1 8 . . . gxf7 19 .ttJgS etc.

b} 9 . . . eS?! has only been seen in one game. It received severe punishment. 1 0. ttJ f5 exf4 I l . ttJxd6t VNxd6 12 . .ixf4 Wicst 1 3 .lt>hl d6 14 .Wif3 0-0 I S . ttJdS ttJg4 1 6.Wig3 ttJceS 17.h3 ttJxd3 1 8 .cxd3 ttJeS 1 9 . .ixeS dxeS 20. WixeS Wic2

2 1 . gxf7! Boom! 1-0 in Mussanti -Triunfetti, Buenos Aires 2002.

Experts vs. the Sicilian 17S

10.�f5 �e7 A tricky move that forces White to

sacrifice a piece. Black has a safer alternative in 1 O . . . !xe3t 1 1 . ttlxe3 d6 ( 1 1 . . .\Wb6?!

12 .\Wd2.) 1 2 .mhl ! with the following split:

a) 1 2 . . . bS? 1 3 .!xbS axbS 14 . ttlxbS \WdS l S . ttl xd6t me7 1 6.eS ttleS 17 .\Wf3 !d7 I S . Eiadl And White has great compensation.

b) 1 2 . . . b6(!) Best according to Ribli. 13 .\We l ! White is planning \Wg3 or \Wh4

1 3 . . . !b7 ( 1 3 . . . 0-0 14 .\Wh4 with a nice attacking position.) 14. ttl c4! and now:

b l ) 1 4 . . . bS I S . ttl xd6t \Wxd6 1 6.eS \We7 17.exf6 gxf6 And now I S .5 is nice for White.

b2) 14 . . . 0-0-0 I S .eS is great for White.

b3) 14 . . . 0-0 I S .\Wh4! EiadS 1 6. Eiae 1 bS ( 1 6 . . . h6 17.eS) 17 . ttlxd6! and White has a huge attack.

c) 1 2 . . . 0-0?! 1 3 .g4 ( 1 3 .Eif3 is also good. The game De Vilder - Kiseleva, Amsterdam 2000, was short and sweet. 1 3 . . . bS 14 . Eih3 ttl b4 I S .eS ttlxd3 1 6.\Wxd3 dxeS 1 7. ttl g4 EidS I S .ttlxf6t gxf6 1 9.\Wxh7t mfS 20. ttl e4 1-0.) 13 . . . dS ( 1 3 . . . bS I4.gS ttl d7 1 S .\WhS! !b7 and now 16 .Eif3 with 17 . Eih3 coming is very good for White according to Timman - I do not see a defence for Black.) 1 4.gS ttl xe4 I S . ttl cxdS exdS 1 6. ttl xdS ttl g3t 17.hxg3 \Wd7 I S .\Wf3 \Wh3t 19 . mgl !g4 20. \W g2 EiadS and White was just

a pawn up in Manso Gil - De la Riva Aguado, Zamora 1 996, even though the game ended in a draw.

1 0 . . . \Wb6?! l 1 . ttl xg7t is a worse version than the text.

l 1 .�xg7t c;t>flI 12 • .L.:c5 YlYxc5t 13.c;t>hl c;t>xg7 14.e5 �e8

The clumsy 14 . . . ttl fgS turned out badly for Black after I S . ttl e4 \Wc6 1 6. ttl d6 5 17.\whS ttl g6 IS .!xf5! exf5 1 9 .ttlxf5t mfS 20. ttl d6 ttlf6 2 1 . \Wh6t me7 22.\Wg7t mdS 23.VMxf6t me? 24.5 EigS 2S .Eif3 \WdS 26.fxg6 1-0 in Abashev - Chernyshov, Voronezh 1 995. 15.�e4 m6

I S . . . \Wc7 1 6.\WhS ttlg6 17. ttl f6!? has been known as good for White since Topalov - Huebner, Wijk aan Zee 1 996. The text move was doing fine for Black, but this game changed theory's verdict from unclear to winning for White. 16.YlYh5 �g6 17JU3 YlYxb2 18.�af1 b6

I S . . . bS should also be answered by

1 9 . ttJ g5! 19.ttJg5!

1 9 . !'l:h3 was unsuccessful after i.b7 20.'lWh6t Wg8 2 1 . ttJg5 !'l:c8 22. ttJxh7 !'l:xc2 23.i.xc2 'lWxc2 24. !'l:gl b5 25.5 'lWxf5 26. !'l:h5 \Mic2 27.h4 'lWf2 0-1 in Paalman - van der Elburg, Dieren 200 1 . 1 9 .• .i.b7 20.gg3 gc8

20 . . . 'lWxa2 2 1 . ttJxf7 Wxf7 22.5 with a winning attack.

21 .�xh7! 2 1 . ttJxf7 Wxf7 22.i.xg6t hxg6

23. 'lW xh8 also wins. 21 ...gxh7 22.gxg6t fxg6 23.�xg6t <bm 24.�xh7 �g7 25..tg6 �b5 26.�h8t <be7 27.�h4t!

Please stay. 27 ... <bm 28.�f6t <bg8 29..tf'7t <bm 30.Axe6t <be8 31 .i.f'7t <bm 1-0.

Game 37 Parligras - Miladinovic Istanbul 2002

1 .e4 c5 2.�a �c6 3.�c3 e6 4.d4 cxc:l4 5.�xd4 �c7 6..te3 a6 7..td3 �f6 8.0-0 h5?!

A speciality of Miladinovic. I have tried this line on several occasions in blitz (it is kind of cool just to lunge forward with the h-pawn in Larsen-style) but I do not believe it is a good move. Too often the h-pawn would just love to get back to h7. And where is the black king going to go? 9.h3 b5

9 . . . ttJxd4 is similar to 8 . . . ttJxd4 except for the position of the h-pawns. 1 0.i.xd4 i.c5 1 1 .i.xc5 \Mixc5 12 . ttJa4 Now with h5 and h3 inserted this move is good.

a) 12 . . . \Mi c6 got Black into trouble surprisingly quickly in Cuartas - Arias, Medellin 2003.


13.c4 d6 14 .!!cl i.d7 I S .li k3 gS 16.'�'f3 me7 17.lLldSt! A typical device. 17 . . . exdS 1 8 .exdS 'lMfb6 19 .E:fel t i.e6 20.i5 1-0.

b) 12 .. . 'lMfc7 13 .c4 d6 14 .E:c1 The black h-pawn is misplaced in this structure. 14 . . . b6 ( I4 . . . id7 went wrong for Black in the following very instructive game. I S .lLlc3 'lMfcS 16.a3 gS Played in the grand style, but very risky! 17 .'IMf f3 'lMfeS 1 8 .'lMfe3 me7 19 . E:fel E:hc8 20. E:cdl h4 2 1 .'lMfd2 '1Mff4 22. E:e3 mfS 23.if1 E:c6 24. lLle2!. Winning material. 24 . . . lLlxe4 (24 . . . 'lMfeS 2S. lLld4) 2S.'lMfel 'lMfxflt 26.'lMfxfl lLlxfl 27.mxfl E:xc4 28.E:xd6 me7 29.E:d2 g4 30.hxg4 E:xg4 3 1 .lLld4 E:f4t 32.lLlf3 E:c8 33.E:d4 1-0 in Kolev - Miezis, Leon 200 1) I S .b4 lLld7 16.E:el ib7 17.if1 E:d8 1 8 .'lMfd4 eS 19 .'lMfe3 E:b8 20.cS! . This is typical: with the h-pawn gone sailing away White can play very aggressively. We have been following S. Petrosian - Stanke, Germany 2002. The game concluded 20 . . . bxcS 2 l .bxcS 'lMfaS 22.c6 '1Mfxa4 23.E:c4 '1MfaS 24.cxd7t mxd7 2S .E:b l E:hc8 26.E:cb4! . A nasty pin. 26 . . . E:c7 27.i.c4 f6 28.'lMff3 E:xc4 29.E:xc4 idS 30.E:xb8 hc4 3 1 .'lMfxhS 'lMfeit 32.mh2 1-0. 10.�xc6!

White almost always put this knight to sleep after b7-bS. lo ... Wixc6 1 l .ie2!

A new idea, which was first played by Kotronias in 200 1 . 1 l ... Wic7

a) l 1 . . .b4 12.eS bxc3 13 .exf6 gxf6 14.bxc3 E:b8 And now instead of the crazy I S .ixhS?! from A. Vouldis - H. Banikas, Athens 200 1 (0- 1 , 42) White should play I S .E:el !? Here is a possible line: I S . . . ie7 16.c4 ib7 17.if3 'lMfxc4 18 .E:bl 'lMfc8 19 .c4 with good compensation for the pawn.

b) l 1 . . .ib7 12 .i.f3! eS (a concession) 13 .igS ( I3 .lLl dS!? might be better.) and White won after 13 . . . i.b4 14.ixf6 i.xc3 I S .ixg7 E:g8 16.bxc3 E:xg7 17 .E:el 'lMfxc3 18 . E:e3 'lMfcS 19 .ixhS 0-0-0 20.i.g4 mb8 2 1 .a4 dS 22.axbS axbS 23. E:b3 E:d6 24. E:ab 1 dxe4 2S.'lMfe2 5 26.ixf5 e3 27. E:xbS E:xg2t 28. mhl E:g7t 29.ie4 1-0 in M. Parligras - A. Botsari, Kavala 2002. 12.ia gbS 13.id4 b4

13 . . . id6 is a rather risky alternative. Roskar - Kukovec, Dobrna 2002 continued: 14 .E:el ! ih2t I s .mhl ieS 16 .lLldS ! Always look out for this move when Black's king is stuck in the centre. 16 . . . 'lMfd6 17.c3! h4 1 8 .'lMfd2 ib7 19 .E:adl lLlh7 and here instead of 20. lLle3 I prefer the more straightforward 20.ixeS! 'lMfxeS 2 1 .lLlb4 E:d8 22.'lMfe3 and White is much better. 14.e5 bxc3 15.exf6 cxb2 16.E!bl!?

A new move. 16.fxg7 ixg7 17.ixg7 E:g8 1 8 .ixb2 E:xb2 was unclear in V. Kotronias - I. Miladinovic, Patras 200 1 . 1 6 ... gxf6 17.ixf6 ggS IS.ixb2

The black king is homeless, so White has the upper hand.


I8 .• J::�b5 18 . . . !b7 is answered by 19 .�d4! .

I9 . .id4 .ib7

20J�xb5?! 20J l:b3! is a big improvement:

20 .. J 'hb3 2 1 .axb3 !xf3 22.'1Wxf3 'lMfxc2 23J l:al And White has a great attacking position. 20 ... axb5 21 .gel .ig7

21 . . .!xf3! 22.'lMfxf3 'lMfc6 is fine for Black. The rest is rather random. 22.J.xg7 gxg7 23.'lMfd4 EfgS 24.h4 g6 25.hb7 Wxb7 26J�e5 gxe5 27.Wxe5 Wc6 28.Wh8t �e7 29.Wxh5 Wxc2 30.Wxb5 d5 31 .Wb4t �e8 32.a4 WeI t 33.�h2 Wc4 34.Wb5t �e7 35.g3 �f6 36.Wxc4 dxc4 37.a5 c3 38.a6 c2 39.a7 eIW 40.a8W �g7 4I.�g2 Wc4 42.W6 6 43.We3 �f7 44.Wd2 We4t 45.�h2 W6 46.We3 Wd5 47.We2 Wd4 48.Wh5t �g7 49.W6 Wb4 50.�h3 Wb5 51 .g4 fxg4t 52.Wxg4t �f6 53. �g3 Wfl 54.W6t �g6 55.h5t �g7

56.Wg4t �f6 57.Wg6t �e5 58.WgSt �d6 59.Wf4t �e7 60.h6 Wgi t 61 .�h4 WhIt 62.�g4 Wdit 63.6 'lMfc2 64.h7 Wc3 65.WgSt �d7 66.Wb5t �c7 67.WgS �d7 68.f4 Wb2 69.We5 Wg2t 70.�h5 WhIt 71 .�g6 Wbit 72.�g7 WgIt 73.�f7 1-0

Game 38 Almasi - Piket Istanbul 2000

1 .e4 c5 2.�6 e6 3.d4 cxd4 4.�xd4 �c6 5.�c3 ••• Wc7 6-Ae3 a6 7 -Ad3 �f6 8.0-0 �xd4

Black is trying to exchange everything and achieve a draw. 9.bd4 .ic5 IOb5 Wxc5 I 1 .�hl!

This is very similar to Hector's play in Game 34. 1 l .•• b5 I2.Wel!?

Looking to both sides of the board. A


nice flexible move. 12 .. .i.b7

After 12 . . . d6 I recommend 12.£3 but 1 3 .f4 is also possible. 13.a

Just following the recipe from Game 35. 13 ... d6 14.a4 b4

14 . . . bxa4 1 5 J ha4 is, as usual, a little better for White. 15.�a2 as 16.c3

1 6 . .tb5t @e7 ( 1 6 . . . .tc6 17.!xc6t �xc6 1 8 .c3 bxc3 1 9 .�xc3 is good for White says Ribli.) 1 7.c3 transposes to the game. 16 ... bxc3 17.i.b5t rtle7

17 . . . !c6 1 8 .�xc3 is a little better for White. 18.�xc3 ghc8

1 8 .. J 3hd8 might be more logical. White is slightly better after 1 9J 3d1 @f8 20J :id3 @g8 2 1 .�d2. 19.9dl rtlm 20.l:�d3

With his good bishop at b5, pressure

against d6 and more activity, White is better. 20 ... d5!?

20 . . J 'k7 was suggested by Ribli, but after 2 1 . �g3 e5 22J 3fd1 gd8 23 .�e1 �b4 24 . � d2 White i s simply winning the d6-pawn. 21 .e5

2 1 .exd5 �xd5 22. �xd5 !xd5 gives nothing. 21. .. �g8?!

Ribli suggest 2 l . . .� e8!? as an improvement for Black, but then 22.f4! planning a kingside attack looks promising. A sample line: 22 . . . �c7 23.�h4 @g8 24. gh3 h6 25 .gg3 @h8 26 . .td3 d4 and now 27.f5! gives White a winning attack (Fritz is happy!) . 22.�e2

22.f4 �h6 is Black's idea. 22 ... �e7 23.�d4?!

23. �h4!? is a better try. Sample line: 23 . . . @g8 24. gc3 �b4 25.�xb4 axb4 26. gxc8t !xc8 27.b3 !a6 28.!xa6 gxa6 29. �d4 and it seems White has the better ending. 23 .. ..tc6?!

23 . . . � c6! equalizes. 24.g4!

Rules out � f5 altogether. 24 ... ,bb5 25.axb5 a4?!

A strange move. 26.�f2 a3? 27.bxa3!

Now White is winning. 27 ... rtle8 28.�e3 ga4 29J�b1

29.f4 is not bad.


29 . • ':!�b8 30.£4 h5 31 .gxh5 Ek4 32J:�gl �d7 33.E:xg7 E:h8 34.£5 E:xh5 35.l:�xf7 exf5 36.e6t �c8 37.�g2 gh4 38.�f3 E:c2t 39.�g3 E:g4t 40.�h3 d441.gxd4 �d5 42.e7! 1-0

Gatne 39 Ponomariov - Sadler Enghien les Bains 1 999

1 .e4 c5 2.�f3 �c6 3.d4 cxd4 4.�xd4 'ilYc7!?

Another way to reach the mainline Taimanov. Black is sidestepping the 5 . lO b5 line. 5.�c3 e6 6.i.e3 a6 7.i.d3 �f6 8.0-0 �e5

Black' s threat is 9 . . . lO fg4. This is by far Black's best line against White's aggressive system. 9.�f3!?

meek 9 .i.e2 is also possible. The text move is an invention of Ponomariov. It is very tricky and White does not have to play the risky £2-f4 as in the mainline. 9 ... �fg4

Black has 7 (!) alternatives. a) 9 . . . i.d6 1 O .lOxe5 i.xe5 1 l .f4 i.xc3

1 2.bxc3 �xc3 1 3 .e5 lOd5 14 .i.d2 �c5t 1 5 . Whl b5 and now 16J 'kl ! with c4 coming looks promising.

b) 9 . . . d6 1 0. lO a4! lO ed7 l 1 .c4 i.e7 12 . Ekl b6 13 .b4 0-0 14 . E:e1 i.b7 1 5 .i.d4 lO g4 1 6.i.b l E:ac8 17.a3 with an interesting position - later won by White in Mamedov - Esplana, Nakhchivan 2003 .

c) 9 . . . lO eg4 1O .i.d2 i.c5 1 1 .�e2 d6 1 2.h3 lO e5 1 3 .lOxe5 dxe5 14 .�f3 0-0 1 5 .i.g5 and White is slightly better.

d) 9 . . . lOxf3t 1O .�xf3;!;. e) 9 . . . i.c5 10 .i.xc5 �xc5 l 1 . lO a4

�a5 12 . lOxe5 �xe5 1 3 . lO b6 E:b8 14 . lO c4 �h5 1 5 .e5 �xdI 16. lO d6t Wf8 17. E:axdl lOe8 1 8 .i.e4 with an obvious advantage for White, Hector - Buhr, Hamburg 200 1 .

f) 9 . . . lO c4 10 .i.xc4 �xc4 l 1 .e5 lO e4 12 . lOxe4 �xe4 13 . E:el �c6 14.i.d4 and White is better.

g) 9 . . . lO g6 10 . lO a4 is good for White. 10.�xe5

White has an alternative in 1 0.i.f4 if he does not like the text move.

1 0.i.f4 i.d6 ( l O . . . i.e7 1 1 . lOxe5 lOxe5 is fairly equal - but not a draw) . 1 1 .i.g3

9.h3 is the mainline, and the rather lOxf3t 12. �xf3 i.xg3 1 3 .hxg3 lO e5

Experts vs. the Sicilian l S I

14 . �h5 is about equal. In Areshchenko - Bryzgalin, St Petersburg 2003, the better player won: 14 . . . d6 1 5 .�e2 b5 16.a3 �b7 17. !'1ad1 0-0 l S . !'1d4 !'1adS 19 .94! ? liJ g6 20.�h2 e5 2Uid2 liJ f4. This position is fine for Black. 22. !'1fd1 liJe6 23.�g3 liJ d4 24.�d3 g6 25 .�e3 '!Wd7 26.f3 '!Wc7 27. 'it>f2 �e7 2S.a4 �c6 29.axb5 axb5 30. liJ e2 liJ e6 3 1 .�b6 '!Wc7 32.�xc7 liJxc7 33.c4 !'1bS 34. liJc3 !'1fcS 35 .b3 b4 36. liJ e2 !'1aS 37.�b 1 liJeS 3S . liJc l !'1cbS 39.g3 !'1a1 ? This rook soon gets into trouble. 40. liJ a2 'it>f8 4 1 . 'it>e3 'it>e7 42. !'1d3 h5 43.gxh5 gxh5 44. 'it>d2 liJ f6 45 . 'it>cl liJxe4 46.fxe4 �xe4 47. 'it>b2. Trapping the unlucky rook. 47 . . . !'1xb 1 t 4S. 'it>xb1 'it>e6 49. 'it>cl �xd3 50. !'1xd3 f5 5 1 . 'it>d2 e4 52. !'1d5 !'1gS 53. liJxb4 !'1xg3 54. liJ c2 !'1g2t 55. 'it>c3 !'1g3t 56. 'it>b4 !'1d3 57. liJ d4t !'1xd4 5S . !'1xd4 h4 59 . 'it>c3 1-0.

10 . . . .txe3 1 0 . . . liJxe5 1 l .f4

a) 1 1 . . .�c5 12 .�xc5 .'!Wxc5t 13 . 'it>h1 i s a little better for White.

b) l 1 . . . liJ c4 is best says the guru CRibli) . 12 .�xc4 �xc4. Now White can choose between the safe 13 .'!Wd3!? �xd3 14.cxd3 b5 1 5 . !'1acl �b7 16. liJ e2 with a tiny edge or go into the jungle with 1 3 .f5!? Black answers 13 . . . �c5! 14 .�f3 b5 (! Ribli) 1 5 .�xc5 �xc5t 16 .'it>h1 �c7 17 . !'1ad1 with attacking chances for White according to Ribli. I think he is right. Sample line: 1 7 . . . 0-0 l S .f6! �e5 19 .�g4 g6 20.'!Wh4 h5 2 1 . liJ d5!?

c) 1 l . . . liJxd3 1 2.cxd3 and White will enjoy a huge initiative after the coming 13 . !'1cl . 1 1 .tyh5! g6

1 l . . .liJxf1 12 .'!Wxf7t 'it>dS 13 .'!Wh5 g6 14 . liJxg6 !'1gS 1 5 . liJ e5 d6 1 6. liJ f7t 'it>d7 17 .�xfl and, even though he has sacrificed an exchange, White is better. 12.tya tyxe5 13.fxe3 tyg7?!

a) 1 3 . . . f6! is clearly a better try. After 14 .'!Wxf6 �xf6 1 5 . !'1xf6 �g7 1 6. !'18! I think White is better but he has to play very energetically, otherwise Black's pair of bishops will start to tell: 1 6 . . . b5 17 . !'1af1 !'1f8 l S . !'1h3 !'1xfl t 19 . 'it>xf1 h6 20.e5 g5 and now instead of the known 2 1 . liJ e4, I like 2 1 .a4! b4 22. liJ e4 'it>e7 23.a5! and White seems to be much better. How is Black going to free himself?

b) 1 3 . . . f5 14.exf5 �d6 1 5 .g3 exf5 16.e4 is much better for White.

c) 1 3 . . . �c5 ! ? 14 .�xf7t 'it>dS 1 5 .'!W8 b5 16 . '!We2! �b7 17 .!'1f7 with a double-

edged position where I prefer White. I4.ti'f4

Planning e4-e5 and lLl e4. I4 ... d6 IS.eS!

Very important - White must attack! IS ••• dxeS

1 5 . . . ti'xe5!? 1 6JWxf7t @d8 17 .@hl , i t i s rather unclear but I prefer White. I6.ti'a4t .td7 I7..tbS 0-0-0

17 . . . i.xb5? 1 8 . lLlxb5 0-0-0 19 .'iNc4t @b8 20.'iNc7 wins for White (Ribli) . IS.gadI 5!

a) 18 . . . i.xb5 19 Ji:xd8t!? 1 9 . @xd8 20. lLlxb5 axb5 2 1 . gd lt @c7 22.'iNa5t @c6 23.a4 i.d6 24.axb5t @d7 25. 'iNb6 wins (Ribli) .

b) 1 8 . . . axb5 19 .'iNa8t @c7 20. lLlxb5t wins. I9.ti'aS .te7 20..ba6

According to Ribli, White is a little better - and has a safe position - after 20.i.xd7t EExd7 2 1 . EExd7 @xd7 22.'iNb6 @c8 23.'iNxe6t @b8 24. lLl d5 .

This does indeed look good for White,

who will have a dominating knight at d5 after c4. 20 .•• .tg5!? 21 .ti'eSt <ttbS 22..txb7!?

Here Ribli recommends the surprising 22.h4!! in ChessBase Magazine. After 22 . . . i.h6 23. 'iNb6 White is much better. This is very complicated though, and I think I would prefer 20.�xd7t!. 22 • • • <ttxb7 23.gd6 J.eS!

Strong defence. 24J3b6t <tta8 2S.�bS gd7 26.gc6 26 ••• <ttbS 27.l3b6t <tta8 2S.h4 J.h6 29.gxe6 l3gS 30.e4 ti'f8 3I .ti'c6t <ttbS 32.�d6 .te3t 33.<tth2 ti'dS 34.g3 ti'e7 3S.ti'bSt ib7 36.ti'xeS he4? 37.ti'bSt <tta8 3S.�xe4?

38 . EExe4! wins. 3S ••• fxe4 39.ti'a4t L.7 40.ti'xe4t ti'b7 41 .<tth3 ti'xe4 42J:�xe4 ggdS 43.l3fel <ttb7 44.a4 ge7 lh-lh

A fantastic game!

Game 40 Yagupov - Khusnullin Tula 1 999

I .e4 eS 2.�a e6 3.�c3 d6 4.d4 cxd4 S.�xd4 �f6 6.f4

Not our usual move order. The position after move 1 1 would normally be reached by the following move order l .e4 c5 2. lLl£3 lLl c6 3.d4 cxd4 4. lLlxd4 e6 5 .lLlc3 'iNc7 6.i.e3 a6 7.i.d3 lLlfG 8.0-0 d6 9.f4 i.e7 1O .'iNf3 0-0 1 1 .@h l . 6 ••• �c6 7.ie3 a6 s.ti'a ti'e7 9.J.d3


J.e7 10.0-0 0-0 I V�hl

In practice Black is getting slaughtered from this position. The reason is simply that White has a very promising kind of classical Scheveningen: The bishop is at d3 and the queen is already active at f3. (In the classical Scheveningen the queen takes the route e1-g3, but here it might skip g3 altogether and go directly to h3) . 1 l •• .J.d7

a) l l . . .e5 Here I like 12. lLl de2! ? e.g. 1 2 . . . lLlb4 ( l2 . . . J.e6 1 3 .f5; 1 2 . . . exf4 1 3 .lLlxf4) 1 3 J'!:acl and 14.a3 is coming.

b) I l . . J !e8 12 J !ae l J.f8 13 .�g3 lLlb4 14.e5 lLld7 1 5 . lLl e4 lLlxd3 1 6.cxd3 dxe5 17.f5!? �a5 1 8 .fxe6 fxe6 1 9 .�f3 And White won in Korneev - Vidarte Morales, Badalona 1 995 after 1 9 . . . lLl f6 2o. lLlxf6t gxf6 2 1 . �xf6 �c7 22. lLl f3 ig7 23.�h4 �d8 24.�h5 �f8 25 . lLlg5 h6 26. lLl £7 �xd3 27.J.xh6 �h7 28. �e3 1-0.

c) l l . . .lLlxd4 1 2.ixd4 b5 1 3 .e5 lLld5 14.exd6 J.xd6 1 5 . lLl xd5 exd5 16 .�xd5

J.e6 17.�h5 g6 1 8 .�h6 f5 19 .�ael and White is a pawn up and went on to win, Mitkov - Bello Filgueira, Burgas 1 998. 12.gael

12.a4 is also not bad: a) 12 . . . b6 13 . �ae 1 e5 14.fxe5 lLlxe5

1 5 .�e2 lLlxd3 16.cxd3 J.e6 17. lLl f5 J.xf5 18 . �xf5 lLld7 1 9. 1Ll d5 �d8 20. �efl and White was better and later won in Emms - Baczinski, Hamburg 1 995 .

b ) 1 2 . . . �ac8 led to another success for Emms: 1 3 .�g3! .Atypical attackingmove: eyeing g7 and preparing e5. 1 3 . . . lLlh5 14 .�h3 g6 1 5 .f5 lLlxd4 16 .J.xd4 J.f6 17 .ixf6 lLlxf6 1 8 .�h4! with an 'autoattack' . The rest was instructive:

1 8 . . . �d8 19 .�f3 exf5 20.exf5 ic6 2 1 . �h3 lLlh5 22.�g4 lLl f6 23.�g5 �e7 24. �e3 �d8 25 .�fl c,;t.g7 26. �h3 �h8 27.�h6t c,;t.g8 28.fxg6 fxg6 29.J.xg6! �c7 30. �g3 1-0. Emms - Naaktgeboren, Hastings 1 995 .

c ) 1 2 . . . lLlxd4 i s extremely dangerous.

One example: 13 .i.xd4 i.c6 14.�g3 b6 1 5 .e5! . The typical attacking move in this line. 1 5 . . . dxe5 1 6.i.xe5 lWb7 1 7.fS! Opening more lines. 1 7 . . . exfS 1 8J'1xfS lLl e8 19 .�h3 i.d6 20.i.xd6 lLlxd6 2 U ' 1d5! . The winning move. 2 1 . .J'1fd8 22.lWxh7t @f8 23J'1el fS 24.i.c4 1-0 Tseshkovsky - Brodsky, Rostov 1 993. A model attacking game by White. 12 ... hS

12 . . . lLl b4 is a bit tricky. In S . Polgar - Benko, Budapest 1 998 White quickly got an attack going. 13 .i.e2 e5 14.fxe5 dxe5 1 5 .lWg3 lLle8 1 6.lLlfS i.xfS 17.exfS f6 1 8 .�h4 lLld6 1 9 .1'!f3!. Black is now defenceless. The finish was nice: 1 9 . . . lLl£7 20.l'!h3 h6 2 1 .i.c4 l'!fc8 22.i.e6 i.f8

23.i.xh6 gh6 24.l'!g3t @h7 25.�xf6 1-0 B.Wg3!

Again the standard attacking move. Black already has to be very careful - and even that might not be enough.

B ... b4 The (overtly) prophylactic 1 3 . . . @h8?

loses to 14 .lLlxc6! i.xc6 1 5 .i.d4 b4 1 6.e5 lLl e8 (16 .. . dxe5 17 .i.xe5 lWa5 1 8 . lLl e4 l'!g8 1 9 .1Llg5 l'!af8 20. lLlxh7 lLlxh7 2 1 . i.xh7 @xh7 22.lWh3t wins - a typical attack in this line.) 1 7. �h3 1-0. Saltaev - Gikas, Katerini 1 993. 14.�ce2 <;I?h8 IS.�xc6 i.xc6 16.id4!

The right square for the bishop. 16 ... gg8

1 6 . . . l'!ad8 17.e5 dxe5 1 8 .i.xe5 �a5 1 9 . 1Ll d4 .id5 20.fS with a nice attack. 17.eS �e4 18.Wh3 dxeS 19.1xeS id6??

Losing. 20.�d4! f5 21 .�xe6 Wf7 22.ixe4 ixe4 23.�g5 Wg6

24.Wxh7t! Oh yes!

24 .. . Wxh7 2S.�f7 mate. 1-0

The transposition to the Scheveningen

Experts vs. the Sicilian l S 5

is very risky and none of the world's top players enter this line as Black.

It is definitely worthwhile to go over the notes in the previous game as they contain a lot of useful attacking ideas.

Game 41 Hector - C Hansen Malmo 2003

1 .e4 c5 2.�f3 �c6 3.d4 cxd4 4.�xd4 e6 5.�c3 a6

This position can, of course, also arise via the Kan: l .e4 c5 2.<�j f3 e6 3.d4 cxd4 4. lLlxd4 a6 5 . lLl c3 lLlc6. 6.J.e3

6. lLlxc6!? bxc6 7.i.d3 is a good alternative. 6 • • • �f6

6 . . . lLlge7 is the realTaimanovvariation. But against 6 .i.e3 this runs into a strong reply, namely 7. lLl b3!

A bit weird as we have struggled to avoid

this move in other lines. However, in this position there is a traffic jam in Black's position: one of Black's lLls is superfluous (They are on the same circuit.) so White avoids the exchange on d4.

a) 7 . . . d6 S .f4 (S . lLl a4!? is also not bad.) S . . . b5 9 . 'WB lLla5 10 . lLlxa5 'Wxa5 1 1 .i.d3 lLl c6 and White is better. A. Sokolov - Moor, Switzerland 2002 continued: 1 2.0-0 i.e7 13 .'Wg3 g6 14. lLl e2 lLl b4 1 5 . lLl d4 i.b7 1 6.a3 lLlxd3 17.cxd3 'WdS l SJhcl i.f6 1 9 .f5 i.e5 20. 'Wh3 gxf5 2 1 . lLlxe6 fxe6 22.'Wh5t rJ;;e7 23.exf5 �gS 24.'Wxhlt �g7 25 .i.g5t 1-0.

b) 7 . . . lLlg6 S .f4 (S. lLla4 is possible again) and then:

b 1 ) S . . . d6 9.g3! preparing h2-h4-h5. b l l ) 9 . . . b5 1 0.h4 i.e7 1 1 .h5 lLlf8

1 2 . 'Wdl i.b7 1 3 .0-0-0 �cS 14.i.h3 lLl a5 1 5 .i.d4 b4 1 6 .lLld5! and White is attacking, Muir - Lalic, British Championship 1 9S9.

b 1 2) 9 . . . i.e7 1O .h4 0-0 1 1 .h5 lLlhS 12 .a4! White is controlling the whole board and has a huge advantage.

b2) S . . . �bS 9 .g3 i.e7 1 0.h4 0-0 1 1 .h5 The march of the h-pawn is always a problem for Black in this line. We are following Reinaldo Castineira - Ortega Hermida, Lanzarote 2003 . The game ended 1 1 . . .lLlhS 12 .'Wd2 f6 1 3 .0-0-0 b5 14. rJ;;b 1 lLl f7 1 5 .g4 h6 1 6. �gl 'Wc7 17.e5 b4 l S . lLl e4 fxe5 1 9 .95 d5 20.gxh6 lLlxh6 2 1 . fxe 5 lLl f5 22. lLlf6t rJ;;hS 23. lLlg4 lLlxe3 24. 'Wxe3 'Wa7 25. 'Wg3 i.g5 26.i.d3 lLl e7 27.h6 g6 2s. lLlf6 i.e3

I S6 The Kan and Taimanov

29 . .txg6 .txgl 30Jhgl liJgS 3 1 ..tf7 1-0.

c) 7 . . . liJ a5? ? S .liJa4! and .tb6 wins something.

d) 7 . . . b5 and now S.a4! is very annoying. S . . . b4 9 .a5 !

d l ) 9 . . . gbS 10. liJ a4 liJxa5 l 1 . liJxa5 �xa5 1 2. liJ b6 �xb6 13 . .txb6 gxb6 14 .�d4 And White won at move 37, Senff - Vanderwaeren, Leuven 2002.

dl) 9 . . . liJxa5? 1 0. liJ a4! and White wins - a very good trick!

d3) 9 . . . d6 was not a success for Black either after 1 O . liJ a4 .tb 7 l 1 . liJ b6 g bs 1 2.f4 �c7 1 3 . .tc4 gdS 14.�e2 liJ bS 1 5 .5 .ixe4 1 6.fxe6 fxe6 17 . liJ d4 1-0 Fressinet - Moor, Zurich 200 1 . 7.g4!?

An aggressive move that fits well with playing the Keres Attack against the Scheveningen. It is a relatively unexplored line and I expect we will see many developments in this line in the coming

years. 7 . .td3 is possible but Black can reach

a reasonable Scheveningen variation with 7 . . . d6. 7 • • .Ab4

a) 7 . . . h6 S . .tg2 �c7 9 .h3 liJxd4 10 .�xd4 e5 1 1 .�b6 �xb6 12 . .txb6 d6 1 3 .0-0-0 ie6 14 .ic7 'it>d7. Now with .tb6 White keeps a small edge. Instead he went 1 5 .ia5 gbS 1 6.f4 b6 17 . .tb4 'it>c7 I S . ghfl exf4 1 9. 9xf4 .te7 20.e5 dxe5 2 1 . ixe7 exf4 22 . .td6t 'it>cs 23 .ixf4 �-�. Hector - Andersson, Sweden 2000.

b) 7 . . . d5 S.g5! liJxe4 9 .liJxe4 dxe4 10 .liJxc6 �xdl t l 1 . gxdl bxc6 12 .ig2 .id7 1 3 . .txe4 and here White prematurely agreed a draw in Hvenekilde - Jensen, Aarhus 1 976. White is slightly better but was apparently peacefully inclined. 8 . .tgl d5

Ribli recommended S . . . h6!? here. 9.g5!

This is almost always the answer to d5 after White has rushed forward with the g-pawn. 9 ... �xe4 10.�xc6! bxc6

1 0 . . . liJxc3 is answered by 1 1 .�d4 and White has a dangerous initiative after 1 1 . . . liJ b5t 12 .�xb4 bxc6 1 3 .0-O-O. 1 l .he4 hc3t

Forced. The ending after 1 1 . . .dxe4 12 . �xdSt 'it>xdS 1 3 .0-0-Ot 'it>e7 14 . liJxe4 is terrible for Black who has the living dead sitting at cS. 12.bxc3 dxe4 13.Wxd8t �xd8


14.0-0-0t �c7 15 . .if4t �b7 16..ie5

Winning back the pawn and keeping the initiative. Opposite coloured bishops benefit the player with the initiative, and we will therefore see a hard struggle for the initiative in the following moves. 16 ... f6

1 6 .. . l':1g8 17.:i':1he l with an undisputed advantage for White. 17.gxf6 gxf6 IS.J.xf6 �U'8 19..ie5 gS?!

19 .. J �xf2! was better. According to Ribli in ChessBase Magazine White still keeps an advantage after 20. Eihgl Ei£7 2 1 . �ge l c5 22. �xe4. 20..ig3 ga5 2Uld4 e5

Getring the problem piece out. 22J:lb4t �a7 23J:ldl .is 24J::ld6

White is clearly better. Usually Jonny Hector has a very bad score against Cun Hansen, but in this game he brings home the full point. 24 .• J::lc8 25.gf6 i.g6 26.ge6 gbS

27..Axe5 gxb4 2S.cxb4 gxa2 29J:lxc6

Pawn and initiative - White is winning. 29 ... �b7 30.gc5 l3a3 31 .i.c3 if! 32.h4! �b6 33.�d2 l3a2 34.h5 �b7 35.id4 ga3 36.c3 gal 37J:lS .ic4 3S.l3g5 ghl 39.gg7t �c6 40J:lxh7 l3h3 41.h6

The pawn decides. 41 ••• gd3t 42.�e1 gh3 43.ghS �b7 44..ie3 i.d3 45.�d2 .ib5 46.gh7t �c6 47.ge7 �d5 4S.h7 .id3 49..id4 1-0.

And with this game I conclude the repertoire against the Taimanov and Kan. I sincerely hope it will give the reader many successes!

1 88 The Accelerated Dragon

The Accelerated Dragon

- By Peter Heine Nielsen

Game 42 Svidler - Tiviakov Chalkidiki 2002

1 .e4 c5 2.�f3 �c6 3.d4 cxd4 4.�xd4 g6 5.c4

White's most solid and, I think, best choice. If instead White tries to transpose to normal Dragon lines then Black has extra options because he has not moved his d-pawn yet, which can be exploited in many possible ways. 5 .. .J.g7 6.�c2

An interesting sideline gaining in popularity these days. I have quite some experience on the Black side of the Maroczy systems, and always felt most uncomfortable when White kept as many minor pieces on the board as possible. It

is rather strange, but to my mind White would prefer either to keep all four minor pieces, or to exchange them all! From the famous game Botvinnik - Toran (see below) we know that this structure with just rooks on is very uncomfortable for Black.

Botvinnik - Toran, Palma de Mallorca 1 967

White is better, the question is whether to take on d5 with the rook or the epawn. 22Jhd5!

22.exd5 Eic7 23. Eide2 Eig7 and despite Black's clumsy rooks, he will bring the king to f8 and slowly reactivate the g7-rook. White is, of course, better but it is difficult to find a way to make serious progress. This is much easier in the game. 22 ... Eic6?

22 . . . Eic7 23.e5 dxe5 24.fxe5 f5 25 . Eiedl


rllg7 is better for White, but Black keeps reasonable drawing chances. 23.e5 dxe5 24.fxe5 :aoo 25. rlln :am 26.:ad7 fxe5t 27.c;t>e3 :ab8 28.c;t>e4 c;t>g8 29.c;t>d5 c;t>f7 30.:axe5 :ad6t 31 .:axd6 exd6 32.c;t>xd6 :ad8t 33.c;t>c7 :ad2 34.c;t>xb7 gxg2 35.c5 :axh2 36.c6 :ac2 37.b4 1-0

A typical Botvinnik move: simple but very strong. This game has become a classic example of how to win with the Maroczy.

So, why not exchange as many pieces as possible and get closer to the goal? Well, Black will not cooperate. He will happily exchange some minor pieces, but will try to keep some on the board as well. Especially, White has to watch out for the scenario where Black ends up with a knight against a white squared bishop.

How often have I had positions like this as Black?

White should avoid such positions at any cost. White sooner or later will have to sacrifice on d4, but he will be hard pressed to make a draw. Of course when seeing this position it is obvious Black is better, however what normally goes wrong is that White realises too late that Black's idea with knight against white squared bishop is e7-e5 ! This weakens d5 but as no white knight is left, who cares? Then Black reroutes the knight to d4, normally via c5-e6-d4. This Diagram is Black's dream; never let it become reality.

Therefore, I can say from experience that although White would like to exchange all the minor pieces, it felt unpleasant when he kept all four on as well! Black would definitely like to exchange one pair of knights. This is seen in all three major Black systems against White's main line. The old mainline was 6 ie3 ttl f6 7 ttl c3 ttl g4 which these days is much less popular, despite Larsen breathing some new life into the system in the 80s. My favourite was always 7 . . . 0-o 8 ie2 d6 9 0-0 id7 10 'lWd2 ttlxd4 1 1 ixd4 ic6 12 f3 a5 . Why not 1 0 ttl c2 here you might ask, as Black is now committed to putting his bishop on d7? Good question. Experts like Tiviakov therefore play 9 . . . ttlxd4 as Black, which normally transposes back to what Black wants. Still, White has caused Black some problems in this system recently, so the real reason I have not recommended the mainline for White is 5 . . . ttlf6 6 ttl c3 d6


7 .ie2 lLlxd4 8 Wlxd4 ig7. A safe and solid system that, for example, the young Russian Malakhov uses to great effect. AIl I mentioned earlier, it is noteworthy that in all the main systems Black happily exchanges one pair of knights. So why let him? It was Boris Gulko who pointed this out to me. An extra pair of knights in the standard positions is definitely to White's advantage. 6 .. . lLlf6 7.lLlc3 d6 8.ie2 lLld7 9.i.d2

Our main line. This defensive looking move prevents Black's .ixc3. Is it really so clear that .ixc3 is a threat? Again, I 'm not too sure. Nigel Short played .ie3 against Tiviakov and he did not take on c3 . If two such experts agree .ixc3 is not dangerous then White should definitely go .ie3. However, compare this to the English opening: l .c4 c5 2. lLl c3 lLl f6 3.g3 d 5 4.cxd5 lLlxd5 5 . .ig2 lLlc7 There one of White's main ideas is 6.'!Mib3 lLl c6 7 . .ixc6t bxc6 8 .W1'a4. He is actually

willing to sacrifice a tempo in order to be able to take on c6. It is not a direct transposition as here Black has used time on lLlf6-d7, still .ixc3 is a serious idea, even used by White players to go for an advantage. My suggestions are based on id2, but avoiding this slightly passive move is worth a punt, especially for players who enjoy unbalanced positions. 9 ••• 0-0 10.0-0 lLlc5 1 1 .M!?

An interesting pawn sacrifice. The alternatives are worse. For example, 1 1 f3 was once the main line. I still do not see anything wrong with my oid recommendation: 1 1 . . . W1'b6 12 * h i W1'xb2 1 3 E:bl .ixc3 which should be fine for Black. 1 1 .•• ixc3

The principled, brave, but probably bad response to the challenge. 12.ixc3 lLlxe4 13.ib2

Again there are huge similarities to the English opening. Without the move b4 included (and with colours reversed) it would be Vaganian - Kasparov. Garry then had to retreat his bishop to e8, but still gained enough positional compensation to draw. Of course an extra tempo is something, however often sacrifices intended to yield positional compensation for a draw as Black are often not enough for an advantage, even with an extra tempo. However, here there is one huge difference:

White gets to put his bishop on the al-h8 diagonal directly. l 1 .b4 not only


grabbed space, it cleared b2 for the bishop. For those who are not impressed with all kinds of talk trying to justify White's compensation, I will just add that Deep Fritz 8 claims White has an edge here despite the pawn minus. 13 ... te6

How to deal with this as Black then? At the time I thought Tiviakov's approach was correct. Try and attack c4 in time. As Svidler effectively refutes this, Black has to look in other directions. An obvious try is to block the al-hS diagonal in time. This makes sense, but White keeps a dangerous initiative.

An instructive game is: 1 3 . . . e5 14 .1l;Ve l ! ? Freeing dl for the rook. The queen is fine on e l , as White intends to push his f-pawns in order to pressurize on the long diagonal. 14 . . . \1Mg5 1 5 . Eldl i.e6 16. �hl ( 1 6.i.d3! ? f5 1 7.f3 ttJ f6 1 8 .f4 seems like an obvious improvement) 1 6 . . . \1Mh4 1 7.i.f3 f5 1 8 .b5 ttJ d8 19 .93 \1Mf6 20. ttJ e3 ttJ f7 2 1 .i.g2 Elac8 22.f3 ttJc5 23.f4 with good compensation and later 1-0, Milos - Spangenberg, Argentina 1 995 .

1 3 . . . \1Mb6 1 4.a3 i.e6 i s a way of trying to get the positive sides of Tiviakov's idea, without facing the rout as in the game. 1 5 .\1Mc1 ! ? f6 1 6.\1Mf4 ttJ g5 17. ttJ e3 was promising for White in Gausel - El Taher, Moscow 1 994. I like the idea of activating the queen before putting the knight on e3, but even the immediate 1 5 ttJ e3 should give excellent

compensation. 14.h5

Ugly but strong. It of course weakens the c5 square, but the fact that it wins tactically is more important. 14 ... �e5?

14 . . . ttJ a5 is given by Svidler as the only chance. He thinks White has excellent chances after 1 5 .\1Md4 ttJ f6 1 6. ttJ e3 \1Mc7 17 . Elac1 \1Mc5 1 8.\1Mh4 Elac8 1 9 . Elfd l . I see no reason to disagree with him on that one.

14 . . . ttJ b8 1 5 .\1Md4 ttJf6 16.g4 \1Mb6 may seem OK for Black, however White keeps a huge initiative even without the queens: 1 7.g5 \1Mxd4 1 8 .i.xd4 ttJ e4 ( 1 8 . . . ttJ fd7 19 .�f3 just wins b7) 1 9 .�f3 d5 20.i.g2 when Black's centre is about to collapse. 15.Wd4 �f6 16.f4 �ed7

1 6 . . . ttJ eg4 17.h3 ttJ h6 1 8 .g4 1eft Black lost in van Wely - Gustafsson, Dieren 1 999. 17.g4!


Wins a piece and thus the game. It is amazing that a top professional like Tiviakov loses like this in a very computerized age. Probably he made the mistake of trusting Catsten Hansen's and my book which recommended 1 3 . . . !e6. 17 • • • Wb6 18.5

The rest is easy. 18 •• .hc4 19.hc4 c!L)xg4 20.J.d5 gac8 21.c!L)e3 c!L)ge5 22.Wxb6 c!L)xb6 23.J.xb7 gb8 24.J.a6 c!L)d3 25.i.d4 c!L)b4 26JUcl c!L)xa6 27.bxa6 e5 28.fxe6 fxe6 29.gabl 1-0

Game 43 Aronian - Vorobiov Aeroflot Open 2004

1 .e4 c5 2.c!L)a c!L)c6 3.d4 cxd4 4.c!L)xd4 g6 5.c4 J.g7 6.c!L)c2 c!L)f6 7.c!L)c3 0-0 8.i.e2 d6 9.0-0 c!L)d7 10.J.d2 c!L)c5 1 1 .b4 c!L)e6

The sane choice. Black tries to establish control over the d4-square, as usual in the

Maroczy hoping to secure it for one of his knights. White's b4 of course grabbed some space, however it has a downside as well. Soon Black will go . . . a5 asking the question: Will White weaken c5 by playing b5, or will he allow the a-file to be opened by answering it with a3? 12.gcl!?

12.:i!bl has been more populat, but this move has its hidden points. Mainly it protects c3, which will soon become important. Also it leaves the a-file thus not allowing Black to exchange his rook there. 12 • • • c!L)ed4

1 2 . . . a5 seems like a more logical move. Why not at least get the a-file opened? The b6-square is weakened, but this does not seem relevant in this exact position. 13 .a3 axb4 14.axb4 ttJ ed4 1 5 . ttJ xd4 ttJxd4 16 .!e3! e5!? ( l 6 . . . ttJxe2t 17JNxe2 !e6 I B J �fdl !xc3!? ( l B . . . llNc7 1 9. ttJ d5 !xd5 20.exd5 gives White an edge. The tactical point is that 20 . . . 1%a4 2 1 c5! is very strong and even after the better 20 . . . b6 2 1 !d4 .if6! 22 llN e3 White has a serious initiative although Black benefits from the fact that the a-line has been opened.) 1 9 . 1%xc3 Wc7!? ( l 9 . . . WbB 20.!f4 1%cB 2 1 . 1%ccl b6 22.e5 dxe5 23.!xe5 llNb7 24. 1%el h5 25 .!d4 Wh7 26.h3 1%c7 27.We5 1%gB 2B .llNf4 was basically winning for White in Speelman - Pigusov, Sochi 1 9B2. It is noteworthy that the presence of opposite coloured bishops makes it much worse for the defender. He cannot oppose on the black squares.) 20.c5 1%fcB


Oust before this book was to go to the press I noticed that the following game had been played: 20 .. J �fdS 2 1 . .if4 �cS 22.h3 dxc5 23.l:'!:xc5 Elxdl t 24. �xd l �dS 25 .�xdSt ElxdS 26.e5 �g7 27 . .ie3 Eld5 2S .Elc l h6 29 . .ic5 .id7 30. Elal f6 3 1 .exf6t Y2-lh. Fressinet -Maze, Val d'lsere 2004. I would not want to defend Black's position in these lines, but there is a drawish tendency you have to acknowledge when you are White against this kind of opening. Besides, 22. �e l ! with the idea of 22 . . . dxc5 23. Elxc5 and Elxd l is not with check looks like an obvious improvement. I would like to make the reader believe that it is my great understanding of chess, and not my ability to press ctrl+3 (enabling the Fritz engine - ed. ) , which found this improvement - but I would not enter such a foolish endeavour. )

2 1 . Elccl (2 1 .h3!? seems like an obvious improvement. It is not clear what Black should do except for dxc5 which however leaves White a tempo up on the game. Maybe 2 1 . . .f6! ? is the most useful, controlling some dark squares. Black might draw this fairly often, but it is an unpleasant task and obviously we are playing for two results only.) 2 1 . . .dxc5 22. Elxc5 �bS 23.h3 Elxc5 24 . .ixc5 �c7 25 .�e3 f6 26.f4 EldS And Malakhov drew this somewhat inferior position as Black against Dominguez at the 2004 WC in Libya.) 1 7. lLl b5 ! ? I like this direct approach

although it has never really worked out in practice. If Black manages quietly to finish his development he should be fine. l 7 . . . lLl xb5 ( 1 7 . . . El a2 1 S . .id3 ( 1 S .El e l ! ? Seems like the obvious improvement, not fearing lLl xe2 and questioning Black's knight on d4 immediately.) 1 S . . . .id7!? was drawn in Geller-Velimirovic, Skara 1 9S0 . Black i s now very active.) 1 S .cxb5 .ie6 was seen in two Geller-Pigusov games. 1 9 . .ic4 �d7 20.�d3 ElfcS was agreed drawn in their first encounter, Sochi 1 9S9 . The second , Cappelle 1 992, went 1 9 .b6!? f5 20.f3 Ela3 2 1 . �d2 El a2 was decent counterplay in the second. My recommendation is 20 .ic4. The point being the positional pawn-sac. 20 . . . .ixc4 2 1 . Elxc4 f4 22.�d5t � hS 23 . .td2 �xb6 24. Elfc l , which to me looks like excellent compensation. 13.tLlxd4 tLlxd4 14 . .ie3!

Again this is the key motiE Here without the a-file open e5 does not make much sense, so Black has to go for lLl xe2. 14 •.. tLlxe2t 1 5.�xe2 b6

The bishop pair is not a major factor here. White can easily exchange the dark squared bishops and Black lacks a way of creating counterplay. White has a huge edge. 16JUdl .ib7 17 . .id4 hd4

Probably the ugly 1 7 . . . f6 needed serious consideration. 18J�xd4 Y!!c7


19.h4! A typical thrust leaving Black with an

unpleasant choice. To allow the pawn to settle on h6 or to weaken himsdfwith h5, allowing an eventual g4 opening lines. 19 .. J!ac8 20.h5 a6 21 .flld5 bd5 22.exd5!

Well, no need to be too dogmatic. Yes, I spoke highly of E:xd5 in such positions. However, here White already has something going on the kingside and Black has no time for the typical e6 break. After 22 E:xd5 a5 would give some counterplay. 22 .. JUeS 23J:�e4 'ilYd7 24Jle1 b5 25.cxb5 axb5 26.h6 i>f8 27.m2 f6 2S.:axe7!

Crashing through. 2S .. . 'ilYxe7 29.:axe7 i>xe7 30.'ilYe2t i>f7 31 .'ilYxb5 :aclt 32.i>h2 l�l:c2

But simultaneously Black resigned. Just pushing his a-pawn wins easily for White. 1-0

Game 44 Bologan - Motylev Togliatti 2003

1 .e4 c5 2.flla g6 3.d4 cxd4 4.fllxd4 fllc6 5.c4 fllf6 6.fllc3 d6 7.fll c2 Ag7 S..te2 0-0 9.0-0 flld7 10.,td2 a5

A logical move, stopping White's space grabbing b4. The drawbacks are the weakening of the b6 and, especially, the b5 square. 1 1 .flla3!?

1 1 . � cl followed by ih6 might also claim a small edge.

1 1 . E:e l lLl c5 12 .if1 b6 13 . lLl a3 ib7 14 .E:cl E:c8 1 5 .ig5 lLl d4 was reasonable for Black in van Wdy- van der Wid, Leeuwarden 2004, but I guess White keeps a tiny edge in a complex position.

l 1 .ie3 . I once had a lot of sympathy for this . The point is that now ixc3 is much less attractive for Black as a5 has weakened squares in the b-line and Black


will not have � a5 pressurizing the c4-pawn. However, White's most natural plan is sooner or later �d4, which will then lead to some standard positions with White having wasted some time. For example, 1 1 . .. � c5 12 . � d4 a4 ( 1 2 . . . �xd4!? 1 3 .�xd4 �d7 gives Black a reasonable version of one of the Maroczy main lines, though White may still have some edge) 1 3 .�db5 i.e6 14J'l:bl 1Mfa5 which actually is a transposition to a later mentioned Dominguez-Malakhov game. 1 1 �e3 has its points, but as what Black wants is to establish himself on the b5 square, why go via d4 allowing Black a desirable exchange? 1 l ... tLlc5 12.tLlab5 tLld4

Black insists on exchanging knights. And why not? White just lost a lot of time going �c2-a3-b5 . However, he has a strong retort prepared. 12 . . . �e6 13 .�e3 a4 14 .�bl (I am not sure why this has to be played, but it is the only move seen in practice and by some very strong players indeed. 14 .�cl !? �a5 1 5 .f4 to me seems logical and strong. As usual in Maroczy positions with all minor pieces still on the board, Black finds it hard to develop naturally. He lacks space.) 14 . . . �a5 1 5 .f4 ( 1 5 .�cl was Morozevich's move, intending to go �d5 at some point without allowing Black to swap queens on d2, which would be the obvious square for the queen. 1 5 . . . �fc8 1 6.f4 �d8 17.�el �b4 1 8 .1Mfd2 with the usual edge for White in Morozevich

- Iskunsnyh,Togliatti 2003, a game later won by Black though!) 1 5 . . . 5 ( 1 5 . . . a3! ? seems to work, which i s one strong argument in favour of 14 �cl ! ? Here the point is that Black seems to survive the tactics after 1 6.5 ( 1 6.e5 axb2 17. �xb2 seems like White's best option. Despite his shattered pawns, White's central pressure gives some hope, at least of equality.) 1 6 . . . axb2 17.fxe6 �xc3.) 1 6.exf5 ( 16. e5!) 16 . . . i.xf5 17. �cl �b4? 1 8 .g4 �d7 1 9 .�d5 1Mfa5 20.�d2 (20. �xe7t! �xe7 2 1 .�xd6 wins outright.) 20 . . . 1Mfd8 2 1 . �c3 with a huge edge in DominguezMalakhov, Tripoli 2004.

If you are not too impressed with the quality of that game, keep in mind it was the deciding 6-5 minute blitz game of their Tripoli 2004 we encounter. Many, including me, have made worse errors in that situation.

12 . . . 5 is a logical and aggressive choice by Black. This is how White tries to fight for an edge with coloured reversed and thus an extra tempo. Here Black might argue that �a3-b5 was indeed a bit slow, and therefore direct action is justified. 1 3 .exf5 i.x5 14.�e3 seems to give White an edge. Black's problem is: what's next? 1 2 . . . 5 certainly compromises his position, but should give some activity in return. Here I do not see how Black can intensifY his pressure, which means White's positional advantages are more weighty. 13.tLlxd4 .hd4 14�h6!?


Forcing a desirable swap. 14 ... .ig7?!

14 . . . ixc3! ? 1 5 .bxc3 Eie8. I think this is the better choice for Black, if only because here he can play for the win too. As usual in ixc3 structures, it however hurts a lot that the black pawn is on a5 . This is actually the only reason that I think White can claim an edge. 1 6.£3 is the correct approach for White. A direct attack with, for example, Wi d4 will not succeed. Black will put pawns on e5 and f6 anyway, no need to force him to do necessary deeds. Now White will put a rook on b l , the queen on d2, play 'it>hl , etc. I think White has a slight edge, but it is a very complex position. If this does not suit you, I would recommend doing Moro's move order 1 1 ie3. 15 . .ixg7 @xg7

16 . .ig4! White's point, without this his play

would make much less sense. Getting

rid of the white squared bishops is huge progress. Often White ends up with that bishop being bad; this is an integral part of Black's counterplay in the Maroczy. So why is the 3 .ib5t system against 2 . . . d6 in the Sicilian not more popular then? It often ends up as a Maroczy with the white squared bishops exchanged. Well, as usual generalizing such concepts is impossible in chess. I guess it is again due to the fact that White prefers four minor pieces on the board rather than three, but will be happy to swap down to one or none. Three or two seems to favour Black somehow! 16 ... .ixg4

16 . . . ie6 17.Wie2 when White eventually will be ready for ixe6 fxe6 e5! leaving him structurally clearly better. 17.YlYxg4 a4 18.YlYe2 YlYa5 19.9ac1

White's edge is bigger than it might seem at first sight. Apart from being solid Black has no plusses. His a-pawn march did not bring much joy, White managed to protect c3 in time, which means the undermining . . . a3 is poindess. White simply has control of the centre and the possibility of playing on the kingside for free. 19 ... �d7 20.�d5 �f6 21 .gfdl �xd5 22Jhd5 YlYa6 23.M

Well, we have been here before. Such heavy-piece middlegames are just much better for White. 23 ... gfc8 24.h5 e6 25JM4 gc5 26.YlYd2

Excellent judgement by Bologan. White still has a huge edge despite the


simplifications. 26 .. J!xh5 27J�xd6 Wa5 28.Wxa5 ghxa5 29J:U7 g5a7 30.e5 h5 31.gd6 gh8 32.£4 1i.>f6 33.lt>fl g5 34.g3 ge7 35.lt>e3

White is winning. The king enters with decisive effect. 35 ... gxf4t 36.gxf4 gg8 37.ge2 ggl 38.e5t 1i.>f5 39.e6 gg3t

39 . . . ge1 t 40. @d4 @xf4 4 1 . @c5 wins, although Black can play on a bit longer than in the game. 40.lt>d4 l:U3 4U�d7 1-0

Game 45 Gulko - Nielsen Esbjerg 2000

l .e4 e5 2.�f3 g6 3.d4 cxd4 4.�xd4 i.g7 5.e4 �c6 6.i.e3

6. � c2 is our preferred move-order 6 ... �f6 7.�c3 0-0 8.J.e2 d6 9.0-0 i.d7 10.�c2

Why discuss this position? Well, �f6-d7 -c5 is not Black's only plan. Although this position arises far more often by the move order in the game rather than via our early � c2, Black can choose to enter this position, only giving us some not too relevant extra options like putting the bishop on g5 instead of e3, etc. As mentioned earlier Black has started going 9 . . . �xd4 to avoid this exact position, certainly most players prefer putting the knight on d7, not the bishop. 10 ... a6

A favourite of Larsen. 1 0 . . . Wa5?! 1 1 .f4 Eiac8 12 .Eib1 a6 1 3 .b4 VNd8 14 .�d3 was clearly better for White in ShortAndersson, Wijk aan Zee 1 990. Black is clearly suffocating, and the weakness of the b6-square makes things even worse. 1 1 .f3 ge8 12.Wd2

12 .Eie l ! ? is an interesting move order. After 1 2 . . . Eie8 Short went back to the game with 13 . � dl, but why 12. Eiel might be a tad more exact was shown by Schlosser against me as after 1 2 . . . � e5? ! 1 3 .�a3 �a5 he had the very unpleasant 14 . Wb3 eyeing the b6 square, and forcing me to retreat with 14 . . . �c6 just to get b4 for my queen. I was definitely suffering in that game. 12 ... ge8 13.gac1 Wa5 14.�a3

14. Eifd1 ! ? is interesting as well. Originally I liked Black due to the game Anand-Larsen, Roquebrune (rapid) 1 992, won by my great compatriot after: 1 4 . . . � e5 1 5 . � a3 h5 16 .@f1 i.a4!?


17 J �el ic6 when Black had decent counterplay. However Shon came up with the space grabbing 1 5 b4! � d8 1 6 tLla3 a5 17 b 5 ie6 1 8 tLla4 tLl fd7 1 9 b6! and held an serious edge against Felgaer, Argentina 200 l . 14 ... Ae6 15.tLlabl

Definitely not as ambitious as Short's approach, but White will potentially expand on the queenside, and in the meantime Black finds it hard to come up with a good plan for counterplay. 15 ... tLle5 16.b3 gbS 17.a3 geeS IS.M Y!YdS 19.tLld5 b6 20JUdl a5 21.h3 axM 22.axb4 Ad7

Again the only positive thing to say about Black's position is that it is solid. The extra set of minor pieces compared to the normal positions definitely favours White, and the weakness of b6 is also a factor. 23.tLla3 tc6 24.!f1 tLled7

A mistake but an instructive one! 25 . tLlc3 followed by tLl ab5 keeps the edge. 25 ... ta4!

Finally I manage to exchange some minor pieces and get decent counterplay. An exchange could have been made earlier on d5, but that would change the pawn-structure in White's favour. Now everything is OK. 26J3el tLlxd5 27.exd5

Well, it is not always bad for Black to take on d5 . Here I will find it easy to protect e7, and will soon be active on the a-line. 27 ... !xc2 2SJ:lxc2 gaS 29.i.g5

With a draw offer. Not even bothering to check if I knew the standard reply 29 . . . if6! intending to take back on f6 with the pawn, ridding myself of the e7 weakness. Thus White would not have taken on f6, but might have gone 30 h4!? with a balanced game. Ih-Ih.

The Scheveningen

- By Viktor Gavrikov

After 1.e4 c5 2.lLlf3 d6 3.d4 cxd4 4.lLlxd4 lLlf6 5.lLlc3 e6 the move 6.g4 was introduced into tournament practice by Paul Keres in his game against Efim Bogoljubow, Salzburg 1 943. The obvious idea is to play g4-g5 and thereby gain space and create attacking possibilities on the kingside.

Black has three main continuations: He may further his own plans with 6 ... a6 or 6 ... lLlc6 and allow g4-g5, or he may discourage the advance of g-pawn by 6 ••.

h6. Against other moves it is easier for White to fight for the initiative: 1) 6 ... d5 7.exd5 lLlxd5 8.i.b5t i.d7 9.lLlxd5 exd5 10.VNe2t i.e7

After 1 0 . . . � e7 l 1 .ie3, both l l . . .g6? !

(Or l l . . .a6 12 .ixd7t 'Llxd7 13. 'Ll f5! ? �e6 14.0-0-0 0-0-0 1 5 . �d3 with advantage for White.) 12 .ixd7t 'Llxd7 13 . 'Llb5 'Ll e5 14.0-0-0 ig7 1 5 J 'hd5 Fischer - Reshevsky, USA (ch) , New York 1 966, and 1 l . . .'Ll c6?! 12 .0-0-0 g6 13 J �!he 1 ig7 14 . 'Llxc6 bxc6 (Rosen - Fronczek, Germany 2000.) 1 5 .id4! �xe2 1 6.ixg7! �e6 17.ixc6 Black loses a pawn without compensation. n.lLlf5 <it>f8 12.hd7 VNxd7 13.i.e3 lLlc6 14.0-0-0 i.f6

14 . . . E!d8 1 5 . E!hel �c7 16 .�bl h6 17 .h4 with a dangerous lllltiatlVe, Bebchuk - Shamkovich, Moscow (ch) 1 964. 15.VNf3 d4

Luther - Rahls, Bad Wildbad 2000, and now instead of 1 6.if4 g6 stronger was: 16.g5 i.e5 17.<it>bl!? Intending to meet 17 .•. gd8 with 18.lLlxd4 i.xd4 1 8 . . . 'Llxd4 1 9 .ixd4 ixd4 20.c3. 19.c3

200 The Scheveningen

Wd5 20.Wxd5 �xd5 21 ..txd4 gxg5 22.i.xa7!

2) 6 ... J.e7 7.gS �fd7 S.h4 8 .!e3 with the idea !xg5?! 9 .!xg5

�xg5 1 0 .lildb5 is interesting. S ... �b6

8 . . . lilc6 transposes to the variation 6 . . . lil c6 7.g5 lild7 8.h4 !e7. 9.J.e3 0-0 10.a3!? d5 n.e5 �c6 12.f4 .id7 13.Wd2 gcs 14.b3 J.c5 15.�ce2 We7 16.a4 with a slightly better position for White, Gruenfeld - Volke, Biel 1 993.

3) 6 . . . e5 7 . .ib5t .id7 S . .ixd7t Wxd7 9.�f5 h5 10.gxh5!

With the following possibilities:

a) 10 ... �xe4?! 1 0 .. J hh5?! 1 1 . lild5 �xf5 12.exf5 lil c6

13 .!e3 and Black does not have enough for the exchange, Tishin - Iljushin, Tula 2002. 1 1 .�xg7t .ixg7 12.�xe4 d5 13.h6!

dxe4 14.Wxd7t <bxd7 After 14 . . . lilxd7 1 5 .hxg7 �g8

1 6. �g1 lilc5 then 17 .!h6 �c8 1 8 .h4 is unpleasant, Aseev - Epishin, USSR 1 988. 15.hxg7 �gS 16.gg1 �c6 17 .J.e3 �e7 IS.0-0-0t <be6 19..ic5 �f5 20 . .if8 �d4 21 .<bbl! <bf5 22.0 �a?!

More stubborn was 22 .. . lil e6 23.!a3 �xg7 24. �xg7 lilxg7 25 . �d7 'ktg6 26. �xb7 �h8. 23.�hl ! ggrlS 24.grlSW gxfll 25.h4 and White won in Bologan - Timofeev, Europe - Tartarstan 200 1 .

b) 10 ... �xh5 n . .ih6! �c6! 1 1 . . .g6?! 1 2.!xfll gxf5 13 .!xd6 lilc6

14 .!c5± . 12.Wxh5 �e7

1 2 . . . g6?! 13 .�g5 gxf5 14 .!xfll 'ktxfll 1 5 .0-0-0 lild4 1 6J :'!:hg1 �e6 allows White an excellent attacking position after 17 . �xd4! exd4 1 8 .�g7t 'kte7 1 9 .1ild5t 'ktd7 20.�xd4 �hc8 2 1 . �b4! �ab8 22.�a4t �c6 23.�xa7 �h6t 24. 'ktb1 , Shmuter - Obukhov, Russia 1 993, or 20 . . . b5 2 1 .exf5 �e5 22.�d3 'ktc6 23. �d1 , Korneev - Suba, Zaragoza 1 995 . 13.�xe7

Less convincing is the recommendation of Nunn: 13 .�g4 lilxf5 14.exf5 �xh6 1 5 .0-0-0 0-0-0 1 6.f4 in view of 1 6 . . . �g6 17.�h3 exf4. 13 .. .he7 14.0-0-0 gxh6 15.We2 0-0-0 16.<bbl <bbS 17.�d5 �dhS IS.a


with slightly better prospects for White, Ye Jiangchuan - Minnebo, Geneva 1 997.

After these minor lines, let us turn to the first main line.

The variation with 6 ••• a6 6 • • • a6

A standard Sicilian move. Black prepares b7-b5 and plans to organize counterplay on the queenside. 7.g5 �fd7 8.h4 b5

If Black plays 8 . . . � c7 9.h5 b5 1 0.a3 ib7 White may reply 1 U�h3!? with the idea g5-g6. 9.a3 .tb7

After 9 . . . lLlb6 1 0 .h5 Black has several possibilities which all seem insufficient to equalize:

a) 1 0 . . . e5? ! l 1 . lLl f5 d5 12 .h6! gxh6 13 .lLlxh6 d4 14. lLl e2 lLlc6 1 5 . lLl g3± Lobron - K. Schulz, Germany 1 985 .

b) 1 0 . . . ie7 1 1 . Elgl (In case of l 1 .�g4

(with the idea l 1 . . .e5? ! 1 2. lLl f5 g6 13 .hxg6 fxg6 14.ie3! gxf5 1 5 .exf5 andWhite has a dangerous initiative for a piece) 1 1 . . . � c7 is playable for Black.) l 1 . . .g6 12 .ie3, T. Ernst - Barash, Gausdal 1 99 1 , with the threat of 1 3 .lLldxb5 axb5 14.�d4.

c) 1O . . . lLl 8d7 I 1 . Elh3 lLlc5 ( 1 1 . . .�c7 1 2.ig2) 12.g6 f6 13 . Elg3;!; , Nunn -Walden, Nottingham 1 983. lO.J.e3

We will now deal with the most important moves, A) lO ••• �b6, B) lO ••• �c5, C) lO ••• .te7 and the main move, D) lO ••• �c6.

A) lO ••• �b6 1 0 . . . d5?! l 1 .exd5 lLlb612 .ig2 lLlxd5 1 3 .lLlxd5 ixd5 14 .�g4 gave White a clear advantage in Sax - Fedder, Plovdiv 1 983. 1 l .h5 �8d7?!

The alternatives are: a) l 1 . . .tDc4?! 1 2 . .txc4 bxc4 13 .We2


V!Jc7 14 .0-0-0 liJd7 1 5 .g6 hxg6 1 6.hxg6 �xhl 17.gxf7t @xf7 I B .�xhl liJf6 1 9.1iJf3! V!Jc6 20.liJg5t @e7 2 1 .�dl followed by £2-f4 was very unpleasant for Black in Luther - Nagendra, Bad Wildbad 1 993.

b) After 1 1 . . .d5 White can play 12 .h6!? (with the idea 12 .. . g6? 1 3 .ttJxe6 fxe6 14 .1l;\fd4) 1 2 . . . gxh61 3.VNh5 VNd6 14 .ih3 e5 1 5 . ttJ f5 VNg6 16 .1l;\fxg6 hxg6 17 .ixb6 gxf5 I B .ixf5 d4 1 9 .ttJd5 ixd5 20.exd5 with advantage.

c) 1 1 . . .ie7 12 .1l;\f g4

1 2 . . . ttJ bd7?! ( 1 2 . . . ttJc6 is an untested alternative which is probably a better choice) 1 3 .g6! f5? ! ( l 3 . . . hxg6 14 . ttJxe6! with the idea 14 . . . fxe6? 1 5 .1l;\fxg6t �ffi 1 6.h6 if6 17.hxg7t ixg7 I B . �xhBt ixhB 19 .ih6t �e7 20.VNh7t �f6 2 l .f4, with the idea 14 . . . 1l;\fcB 1 5 .ttJxg7t �ffi 1 6.id4 if6 17.ixf6 ttJxf6 I B .1l;\ff4 followed by h5-h6) 14 .1l;\fg3! ttJffi 1 5 .h6 if6 1 6.hxg7 ixg7 17.gxh7 if6?!

I B . �h6 with a decisive advantage in Yagupov - Yezersky, St Petersburg 1 993. 12.g6!

The strongest continuation. This excellent idea was introduced into tournament practice in the game Adams - C. Hansen, Wijk aan Zee 1 99 1 . Less dangerous for Black is 1 2. �gl g6 or 12 . �h3 d5. 12 ... hxg6 13.hxg6 gxhl 14.gxf'7t @xf7

14 . . . �e7? is refuted by 1 5 .VNg4! VNcB 16 .1l;\fxe6t �dB 17 .1l;\feBt �c7 I B . ttJ e6t �c6?! (Black is also in a hopeless situation after I B . . . �bB 1 9 .ttJxf8 ttJxffi 20.ixb6 ic6 2 1 . VNe7 ttJd7 22.idB �hB 23.1l;\fxd6t �b7 24.ie7) 1 9. ttJ d5! �xfl t 20. �xf1 ttJxd5 2 1 .exd5t �xd5 22. ttJxf8! VNxeB (22 . . . ttJxf8 23. �dl t �c4 24. �d4t or 22 . . . ttJ e5 23. �d1 t �e4 24J �d4t �f5 25 . �f4t, Kotronias

- Kr. Georgiev, Karditsa 1 994, leads to mate and 22 . . . �c6 loses to 23.VNe4t


rJlc7 24. ttJ e6t rJlbB 25 .Wlf4, Forster - Weigler, Switzerland 1 993.) 23.fxeB� gxeB 24. ttJxd7 rJlc6 25 . ttJb6 gxe3 26.fxe3 rJlxb6 27. gdl rJlc7 2B. gd4 with a won ending for White. 15.Y!YSt Y!Yf6

Unsatisfactory is 1 5 . . . ttJf6 16 .Wlxhl Wlc7 17. ttJ f3!. For example 17 . . . rJlgB I B . ttJ g5 geB 1 9 .0-0-0 d5? 20.�h3! icB 2 1 .ixb5! and Black cannot parry the threat of ghl , Svidler - Nepomniashy, St Petersburg 1 996. 16.Y!Yxhl �c5

In the aforementioned gameAdams- C. Hansen Black tried 16 . . . ttJ e5? ! 17 . ttJxe6! Wlxe6 I B .,ixb6 d5. Play continued 1 9 .id4 ttJ c6 ( l 9 . . . dxe4? 20.�h5t ttJg6 2 1 . ih3.) 20.ih3 WleB 2 1 .0-0-0 ttJxd4 22. gxd4 �e5 23. gd3 d4 24. ttJd5 ,ixd5 25.exd5 with good winning chances for White. 17.Y!Yh5t!? g6 18.Y!Yg4 and it is hard to see what Black can do

against the logical follow-up 0-0-0 and ttJf3.

B) 10 . . . �c5 1 l .Y!Yg4!? �bd7 12.h5 This is more accurate than

1 2.0-0-0 Wla5! 1 3 .h5 b4 14.axb4 �xb4 with strong counterplay for Black. 12 ••. �e5 13.Y!Yg2 �c4

1 3 . . . gcB also looks good for White, e.g. 1 4.f4 ttJ c6 1 5 .g6 hxg6 1 6.hxg6 gxhl 17.gxf7t rJlxf7 I B .Wlxhl �f6 1 9.0-0-0 ttJxd4 20. gxd4 ie7 2 1 . ig2;!;, De Vreugt

- G. Horvath, Vlissingen 1 997.

14.hc4 bxc4 15.0-0-0 gb8

The alternative is 1 5 . . . �b6 when 1 6.g6 gbB 1 7.gxf7t rJlxf7 IB . ttJf3! should be dangerous for Black. For example: I B . . . ic6 (if I B . . . iaB, then 1 9. ttJ a4! �b5 20.ixc5 Wlxa4 2 1 .�g4!, or 20 . . . dxc5 2 1 . ttJ e5t rJleB 22.�g4!) 1 9.,ixc5 Wlxc5 20. ttJg5t rJle7 2 1 . Wlg4 id7 22.Wlf4 ieB (after 22 . . . �b6 winning is 23. ttJ f7 ggB 24. ttJxd6 �xb2t 25. rJld2 gb3 26. gh3) 23.e5! d5 (both 23 . . . dxe5 24.�h4 �b6 25 . ttJxe6t! rJlxe6 26.Wlg4t rJlf7 27.�xc4t Wle6 2B.�c7t ie7 29.�xbB and 23 . . . Wlxe5 24.Wlh4 �f5 25 . ghel rJld7 (25 . . . e5 26. gxe5t! �xe5 27. ttJ f3t Wlf6 2B. ttJd5t and 26 . . . dxe5 27. ttJxh7t rJle6 2B.�xc4t rJle7 29. ttJd5t) 26. ttJxe6 are bad for Black) 24. ttJxe6!! rJlxe6 25 .ttJxd5 rJld7 (25 . . . ia4 26.h6! �c6 27. gh5) 26. gd4! �a5 27.e6t rJlcB 2B. gxc4t rJlb7 29. gc7t rJlaB 30.�f3 gb7 3 1 . gcBt gbB 32. gxeB gxeB 33. ttJb4t winning

(analysis) . 16.�de2!?

Now �b6 can be met by iDa4. However, also possible is 1 6.g6 �c7 17.gxf7t �xf7 1 S .�g4 and White has at least a small edge. 16 .•• tya5 17.g6

With a strong initiative for White. Now the game Kulaots - Yewdokimov, Tallinn 200 1 continued: 17 ••• .te7

17 . . . f5 I S .gxh7 iDxe4 19 .�g6t <;!;>d7 20.h6! gxh6 2 1 . �f7t lie7 22Jhh6. 18.grlit c;t>xf7 19.tySt .if6?!

19 . . . <;!;>eS was better, when White plays 20.�g4 with the idea 20 . . . <;!;>f7 2 1 . iD f4 lics 22. iD g6! . 20Jhd6 with a clear advantage.

C) 10 •• ..ie7 1 l .tyd2 Interesting is 1 1 . � e2 iD c6 ( 1 1 . . . iD b6!?)

1 2.0-0-0 0-0 1 3 .f4 iDxd4 14 .lixd4 �a5 , Baikov - Antkowiak, Nuremberg 1 9S9, and now 1 5 .g6!? fxg6 ( 1 5 . . . hxg6? 1 6.h5) 16 .�g4 �f7 17 .�xe6 with better prospects for White.

The continuation 1 1 . �g4 iDc6 12 .0-0-0 gives Black sufficient counterplay after 1 2 . . . iDce5 1 3 .�g3 �cS 14 .f4 iDc4 1 5 .lixc4 �xc4 1 6.f5 iDc5 ( 1 6 . . . e5!?) 1 7.g6 fxg6!, Zhao - Alcazar Jimenez, Oropesa del Mar 2000. 1 l • • • �c6

1 1 . . . iD c5 12.f3 �c7 is less accurate since after 1 3 .0-0-0 White may meet 1 3 . . . iD bd7 (Or 1 3 . . . iDc6 14 .lixb5!?)

with a typical piece sacrifice 14 .lixb5 ! ? axb5 1 5 . iD dxb5 �c6 16 .iDxd6t lixd6 17. �xd6±, Fischer - Najdorf, Leipzig (01) 1 960.

Another idea is 1 1 . . .iD b6!? planning iD Sd7-e5 . Karklins - Commons, Lone Pine 1 972, now continued 12 .iDxe6!? fxe6 13 .�d4 iDc4 (unclear is 13 . . . iD a4 14 .�xg7 �f8 1 5 .�xh7 iDxc3 1 6.bxc3 �a5 17.lid2) 14 .�xg7 <;!;>d7 1 5 .lixc4 bxc4 1 6.0-0-0 �f8 ( 1 6 . . . �gS 17. �xd6t! <;!;>xd6 I S . �d lt lid5 19 .�d4 iDc6 20.�xc4� or 1 6 . . . iDc6 17 . iDa4 �f8 IS .�c3;!; .) 17 . �xd6t!? ( 17 .�d4!? iDc6 I S .�xc4�) 17 . . . <;!;>xd6 I S .�d4t <;!;>c7 19 .�b6t <;!;>cS 20.�xe6t iD d7 (20 . . . <;!;>dS? 2 1 . �dl t <;!;>eS 22. iDd5+-) 2 1 . �dl �dS?? (This loses. Correct was 2 1 . . .�eS! with the idea 22. iD a4? lic5 ! And if 22.�xc4t <;!;>bS 23.lif4t <;!;>a7 24.fie3t, then 24 . . . <;!;>bS 25 .lif4t with perpetual check.) 22. iDa4! and Black has no defence against �xd7. 12.0-0-0 0-0

In Campora - Landenbergue, Biel 1 99 1 , Black tried 12 . . . �cS 13 . <;!;>bl �a5 14.h5 b4?! 1 5 .axb4 �xb4, but after 16 .iDxc6 lixc6 17 .�d4! �xd4 I s .lixd4 lixg5 White could play 19 .1ixa6 �aS 20.lib5 lixb5 2 1 . iDxb5 when Black loses a pawn without compensation. 13.c;t>bl gb8 14.£4 ge8 15.g6!? hxg6 16.h5 �xd4, Vogt - Summermatter,

Lenk 1 990, and here

17.Axd4 gives White good attacking prospects.

Now let us turn to our main game after 1.e4 c5 2.<��f3 d6 3.d4 cx:d4 4.c!Llxd4 c!Llf6 5.tik3 e6 6.g4 a6 7.g5 c!Llfd7 8.h4 b5 9.a3 .ib7 10 . .ie3 with D) 10 ... c!Llc6.

Game 46 Renet - Summermatter Swiss Championship 1 992

D) 10 ... c!Llc6 1 l .h5 1 1 .�e2 lLlde5 ( l l . . . Ek8 1 2.0-0-0

lLlce5 1 3 .f4 lLlc4 is also good for Black, Heinemann - Hetey, Germany 1 998.) 12.0-0-0 lLl c4 1 3 .lLlxc6 ixc6 14.f4 �a5 1 5 . lLl d5 (Alexander - Lundholm, corr. 1 970-7 1 ) was recommended by Nunn in his book with Gallagher Beating the Sicilian 3, but after 1 5 . . . ib7! White has

serious problems! 1 l ... c!Llxd4

The alternatives are: a) 1 1 . . .!e7 1 2.l:'i:g1 g6 1 3 .l:'i:h l !? l:'i:f8

( l 3 . . . hg5 14.hxg6 ixe3 1 5 .gxf7t �xf7 1 6.%Vh5t �e7 17.fxe3 is slightly better for White.) 14 .f4 c!Llxd4 1 5 .%Vxd4 e5 1 6.%Vd2 exf4 17.hf4 tLlc5 ( l 7 . . . tLle5 1 8 .0-0-0 with the idea 1 8 . . . c!Llf3? 19 .%Ve3 tLlxg5 20.tLlxb5!) 1 8 .0-0-0! tLlxe4? 1 9.tLlxe4 he4 20.!g2 hg2 2 1 .%Vxg2 l:'i:c8 22.l:'i:he 1 with decisive threats in Leenhouts - Muhren, Dieren 2003.

b) 1 1 . .Jk8 1 2J'l:h3 lLlxd4 ( l 2 . . . lLl ce5 1 3 .g6! , Ghinda - Tsarouhas, Ano Liosia 1 998, with the idea 1 3 . . . hxg6 14.hxg6 lLlxg6 1 5 . Eixh8 lLlxh8 1 6.'�h5 lLl g6 17. lLlxe6) 1 3 .'�xd4 lLl e5 14 .0-0-0 lLl c4 1 5 .g6 hxg6 1 6.hxg6 Eixh3 17.ixh3 lLlxe3 1 8 .'�xe3 with a dangerous initiative, Tolnai - Janetschek, Austria 1 993. 12.Wxd4 c!Lle5 13.0-0-0 c!Llc6 14.Wd2 gc8 15Jfh3!

This is a typical idea for such positions. White intends to play g5-g6. 15 ... b4 16.axb4 c!Llxb4 17.�b1 Wa5 18.f3 .ie7

1 8 . . . d5 looks risky for Black after 1 9 .96. For example 1 9 . . . e5? 20.exd5 ixd5 2 1 . ib6! %Vxb6 22. lLlxd5 %Vc6 23.ic4! with the threat of lLl b6. 19.96 O-O?

An illogical decision that allows White to organize an attack against the black king. After 1 9 . . . fxg6 20.hxg6 h6 White


has only a small advantage.

20.gg3! An important resource. Now Black has

to reckon with the manoeuvre ifl -h3. 20 ••• 1£6

The other moves are worse: 20 . . . f5? 2 1 .h6 and 20 . . .'IWxh5?

2 1 .gxh7t V;Vxh7 22.ig2. Also 20 . . . fxg6? 2 1 .ih3 is no joyride.

21 '!d4 e5 Black cannot exchange the dark

square bishops: 2 l . . .ixd4? 22.V;Vxd4 e5 23.V;Vxd6 EifdB loses to 24.gxf7t @hB 25 .h6! or 24 . . . @xf7 25.ic4t! . 22..te3?

The natural reply, but this retreat was only strong after the inclusion of the moves 22.gxh7t! @xh7. 22 •.• d5?

Missing a chance by 22 . . . hxg6! 23.hxg6 d5! to punish White for his mistake. 23.gxh7t �xh7

23 . . . @hB would also be met by

24.ig5! . 24..tg5! d4?

Resignation. The only possibility to continue the fight was 24 . . . Eixc3 25 .V;Vxc3 V;Va2t 26. @c1 'lWa4 27.V;Vb3 V;Vxb3 2B.cxb3 ixg5t 29. Eixg5 dxe4 (29 . . . EicBt 30. @bl dxe4 3 1 .ih3 EibB 32.if5t @h6 33. Eig4.) 30.fxe4 EicBt 3 1 .ic4 ixe4 when it is not so easy for White to realize his material advantage. 25 . .hf6 gxf6 26.!h3 1-0

The variation with 6 ••• �c6

7.gS �d7 S..te3 A flexible move that prepares long

castling and, in comparison with B .h4, reserves the attacking manoeuvre Eiglg3-h3.

The main alternative is B . tiJ db5 tiJ b6 9 .if4 tiJe5 (after 9 . . . e5 10 .ie3 ie6


l 1 . lZldS ixdS 1 2.exdS lZl e7 13.c4 White has a small plus) 10 . �hS lZlg6 (worse is 1 0 . . . id7 because of l 1 .ixeS dxeS ( I I . . . g6? 12 .ixd6!) 1 2.g6! a6 1 3 .gxf7t rlle7 14 . lZla3 �c7 I S . �gl !?) l 1 .ig3 (to unclear play leads l 1 .ie3 a6 1 2. lZl d4 dS) I 1 . . .a6 1 2. lZl d4 ie7 1 3 .0-0-0! ixgSt 14. rllb l 0-0 I S .ixd6 �xd6 16 .WlxgS h6 17. � e3 with a slightly better position for White according to Kasparov. 8 •• .i.e7

A central strike is the recommended reaction to a wing attack, but here the plan with d6-dS does not promise Black adequate counterplay:

a) 8 . . . lZlb6 9.h4 dS 10 .ibS id7 l 1 .exdS exdS ( I I . . .lZl xdS 1 2. lZl xdS exdS 13 .�e2 ie7 14.0-0-0) 12 .Wle2 ie7 1 3 .0-0-0 0-0 14 . lZlb3! ie6 I S .f4 �e8 ( l S . . . ib4 1 6.�d3 is similar.) 1 6.hS ib4 17 .�d3 lZlc4 1 8 .icS! ixcS?! (better was 1 8 . . . �b8) 1 9 .1ZlxcS �d6?? 20.�xc4 1 -0, Glek - Dydyshko, Azov 1 99 1 .

b) Sometimes Black tries to create counterplay on the queenside by 8 . . . a6 9.h4 �c7, when White can choose between two queen moves:

b l ) 10 .�e2bS l 1 . lZlxc6! ? Wlxc6 12 .id4 ib7 ( l 2 . . . b4 1 3 .lZldS as 14 .WlbS! WlxbS I S . lZl c7t rlld8 16 .lZlxbS ib7 17.ig2 is favourable for White according to Nunn) 1 3 .0-0-0 0-0-0 14.a3 lZl b6 I S . �h3 rllb8 1 6.f4 with better prospects, Korsunsky - Eingorn, Baku 1 979.

b2) 10 . Wld2 bS 1 1 .0-0-0 ib7 12. lZlxc6 Wlxc6 1 3 .a3 lZl eS 14 .f4! lZl c4 I S .ixc4

Wlxc4 1 6.id4 �g8 17. �h3 ie7 1 8. rllb l �c8 1 9 .f5 with the initiative, Ermenkov - Kr. Georgiev, Bulgaria 1 987.

Game 47 Timmermans - de Jonghe Belgium 1 989

9J�gl Another interesting attacking

possibility is 9 .h4 0-0 1 0. �hS. The game Movsesian - Cvitan, Germany, 1 997 is a good illustration of this idea: 1 0 . . . a6 1 1 .0-0-0 lZlxd4 12 .ixd4 bS 13 .id3! ? lZl eS (I3 . . . b4? loses in view of the typical sacrifice 14 .ixg7! with the idea 14 . . . rllxg7 I S .�h6t rllh8 16.eS etc. The same goes for 1 3 . . . �e8? 14 .ixg7! rllxg7 I S .Wlh6t rllh8 1 6.eS lZlffi 17. lZl e4 lZlg6 1 8 . lZl f6 ixf6 1 9 .9xf6 �g8 20.hS �ffi 2 1 .�xh7t! 1 -0, Madl - Summermatter, Geneva 1 988.) 1 4.f4 lZlxd3t I S . �xd3 ib7? ( l S . . . �e8?! gave White a crushing attack in Naiditsch - Bischoff, 2000 after

16.f5 ifS?! ( l 6 . . . exf5 17Jl: f1 ) 17.f6 g6 1 8 . �f3 e5 1 9 .h5! Black should have played 1 5 . . . b4 1 6. lt.'l e2 �a5 17. Wb1 e5 1 8 .ie3 ie6 1 9 .b3 d5, when White can still claim an edge by 20.�f3! f5 2 1 .gxf6 ixf6 22.f5 dxe4 23. �xe4 if7 24.ig5, Z. Almasi - Cvitan, Budapest 1 99 1 .) 1 6J'l:gl ! (the opposition with the black king is a decisive factor in this position) 16 . . . b4 17. lt.'l d5! exd5

1 8 . �dg3! ! (with the terrible threat �h6) 1 8 . . . �c7 ( 1 8 . . . �e8 19 .96 fxg6 20. �xg6 hxg6 2 1 .�xg6 !f6 22.!xf6 �e7 23.ixg7+-) 19 .�h6! �xc2t 20. <;hc2 �fc8t 2 1 . Wd2 gxh6 22.gxh6t ig5 23. �xg5t WfS 24.exd5 We7 25 .�f5 �c4 (or 25 . . . �fS 26. �g7 ic8 27.!f6t We8 28 .� fg5 winning) 26. Wd3 �ac8 27. �g7 1 -0 . 9 ... 0-0

If Black plays 9 . . . a6 White may reply 10 .�h5 when Black has nothing better than 1 0 . . . 0-0 transposing back into the

main line. 10.Wh5

This looks natural and strong. The space advantage allows White to prepare his forces for an attack on the kingside. 10 .. Jie8

Black vacates the fS-square for the knight and prepares for the manoeuvre �g3-h3 . The alternatives are:

a) 1 0 . . . a6 1 1 .0-0-0 �e8 12 . �g3 (In Akopian - Brenninkmeijer, Groningen 1 99 1 , White included first the moves 1 2.f4 !fS and only now played 13 . �g3 g6 14 .�h4 h5?! 1 5 .!e2 !g7 16 .lt.'lxc6! bxc6 17. �xd6 ixc3 1 8 .bxc3 'l&a5 1 9. Wd2! with a clear edge.) 12 . . . lt.'lxd4 1 3 .ixd4 b5 14 . Wb 1 ! (preventing .!xg5 landing with check after �f3 or �h3) . Now we have three options:

a l ) 14 . . . b4? is bad on account of 1 5 . �f3! f6 ( 1 5 . . . �fS? 16 . �h3 h6 17.gxh6) 1 6. �h3 It.'l fS 17.gxf6 ixf6 18 .ixf6 gxf6 19 . �g3t Wh8 (Or 1 9 . . . lt.'lg6 20. �xg6t

hxg6 2 1 . VNxgGt <;!(fS 22.'!Mfh6t <;!(f7 23 . .te2! with decisive threats .) 20 . VNf7 tLlg6 2 1 . :B:h3 tLlfS 22.�b5! '

a2) 14 . . . tLlfS? 1 5 .:B:dd3! b4 16 . .txg7! <;!(xg7 17. :B:d8 bxc3 ( 1 7 . . . tLlg6 18 . :B:xf7t <;!(xf7 19 . VNxh7t <;!(fS 20. :B:f3t) 1 8 . :B:xf7t <;!(g8 1 9 .9G winning, Zaichik - Siekanski, Polanica Zdroj 1 989.

a3) Probably Black should settle for the worse position after 14 . . . e5 1 5 . .te3 g6 16 .VNh6 �fS 17.'�·h4, as he did in the game Santo Roman - Bischoff, Uzes 1 990.

b) 10 . . . g6 1 1 . 'lMfh6 (unclear is 1 1 . 'lMfh4 tLlde5 12 . .te2 tLlxd4 13 . .txd4 fG, Wohl - Cvitan, Groningen 1 997.) l 1 . . .tLlde5 (the continuation 1 1 . . . :B:e8?! 12 .0-0-0 .tfS 13 .VNh4 a6 14 . :B:g3 .tg7 1 5 .f4 tLlfS? 16.e5! looks very dangerous for Black, Ivanovic - Mascarinas, Manila (izt) 1 990) 12.0-0-0 fG 1 3 .gxf6 �xf6 14 .tLlxc6 bxc6 1 5 .�e2 with slightly better prospects for White in Kengis - Murugan, Gausdal 199 1 .

c) 1 0 . . . tLlxd4 1 1 .�xd4 tLl e5 12 . :B:g3!? (an idea which deserves attention is 12.0-O-O!? g6 1 3 . 'lMfh6 fG 14.gxfG .txf6 15 . .te2 �g7 1 6.VNd2) 12 ... gG 13 .'lMfe2 .txg5 14.�e3 �f6 1 5 .0-0-0 with compensation for the pawn, Jansa - Nielsen, Gausdal 1 990. 1 1 .0-0-0 ttlm 12.f4

Also possible is 12 . :B:g3 planning to meet 12 . . . a6 by 1 3 .tLlxc6 bxc6 14.e5 VNc7 1 5 . tLl e4! dxe5 1 6. tLl fGt gxf6 17.gxf6t tLlg6 1 8 .fxe7 :B:xe7 19 .�g5

with a dangerous initiative, Atri Sangari - Kelly, Moscow (01) 1 994. 12 ••. a6

It is easy to criticise this move, but Black has serious problems since 1 2 . . . tLlxd4 1 3 . .txd4 .td7 14 .�d3 e5?! fails to 1 5 .fxe5 dxe5 16 .�xe5 �c5 17 .�c4!. Better is 14 . . . :B:c8 with the idea of a future sacrifice on c3, when White's chances would only be slightly preferable. After 14 . . . :B:c8, the move 1 5 .f5 with interesting prospects for White suggests itself. 13.gg3 ttlxd4 14..hd4

Thanks to Black's last move the white bishop is able to take up a menacing position. 14 ••• b5 15.i.d3 b4?

Provoking the following combination. The only chance for Black to prolong his resistance was a pawn sacrifice by 1 5 . . . e5 1 6.fxe5 dxe5 17 . .txe5 VNa5 .

16.Lg7! cbxg7 After 1 6 . . . bxc3 White wins with

2 1 0 The Scheveningen

17.'�'h6 �b6 1 8 . .ixc3 e5 1 9 . .ixe5. 17.YlYh6t c;t>g8

Or 17 . . . �h8 1 8 .e5 followed by .ixh7. 18.e5 tLJg6

If 1 8 . . . bxc3 then 1 9 . .ixh7t lLlxh7 20.g6 is decisive. 19.ixg6 fxg6 20.!ih3 .bg5 21 .YlYxh7t c;t>f8 22.YlYh8t 1-0

The variation with 6 • • • h6

6 • • • h6

Black tries to keep his knight on f6 and gain some time for counterplay. 7.h4

For a long time White mostly played 7.g5 hxg5 8 . .ixg5, but now 7.h4 is considered more promising. White wants to continue with ggl (or .ig2) and g5 , driving the black knight away. 7 ... tLJc6

The most natural reply, but Black has

two other possibilities: a) 7 . . . a6 8 .!g2 lLl c6 (8 . . . g6 9 .g5 hxg5

1 0 . .ixg5 .ie7 1 1 .�d2 e5 1 2. lLl de2 .ie6 1 3 .0-0-0 lLlbd7 14.f4 �a5 1 5 .�b 1 is a little better for White.) 9 .g5 hxg5 1 0.hxg5 gxh1 t 1 1 ..ixh1 lLld7 1 2.f4 V;Vb6 ( 12 . . . g6 1 3 . .ie3 �b6 14.a3 gb8 , Fernandez Garcia - Suba, Seville 1 994, 1 5 .V;Vd2 with the idea 0-0-0) 1 3 .lLlde2 g6 14.b3 �c7 (after 14 .. . V;Vc5 White plays 1 5 .�d2 b5 1 6 . .ib2 .ib7 17.0-0-0 0-0-0 1 8 . �b1 �fL. 19 .ge l ! followed by lLld1 -e3, Short - Kindermann, Dortmund 1 986) 1 5 . .ib2 b5 1 6.a3 lLlc5 17.V;Vd2 .ib7 1 8 .0-0-0 b4 1 9.axb4 lLlxb4 20. �b1 V;Va5 2 1 .f5 ! with the initiative, van der Weide - Agdestein, Cap pelle 200 1 .

b) 7 . . . !e7 8 .V;Vf3

8 . . . lLlc6 (if 8 . . . h5 9.gxh5 lLlxh5, then 10 .!e3! ? lLlc6 1 1 .0-0-0 a6 1 2 . .ie2 .id7 13 . V;V g2 looks good for White, Brodsky - Krzywicki, Warsaw 1 993.) 9 .lLlxc6 {dubious is 9 . .ib5?! because of 9 . . . .id7


1O .i.xc6 bxc6 1 1 .g5 hxg5 l 2.hxg5 gxhl t 13 .�xhl tLlg4! 14 .�h3 �b6 1 5 .�xg4 �xd4, Agopov - Gavrikov, Helsinki 2000.) 9 . . . bxc6 10 .g5 tLl d7 I I .gxh6 gxh6 12.i.d2!? i.xh4 1 3 .0-0-0 �e7 14 .�f4 i.f6 ( l4 . . . i.g5 1 5 .�g3) 1 5 . gxh6 i.b7 16J �xh8t i.xh8 17.�g3 0-0-0 1 8 .i.g5 i.f6 19 .i.xf6 tLlxf6 20.e5 dxe5 2 1 . gxd8t �xd8 22. � xe5 with a favourable ending for White in Gallagher - Cu. Hansen, Germany 2002. S.ggl

Now Black has a choice between two main continuations A) S ••• dS and B) S •.• hS. Possibilities like 8 . . . g6 9.g5 hxg5 1O .i.xg5 i.e7 1 1 . �d2 a6 l 2.0-0-0 or 8 . . . tLld7 9.g5 hxg5 1O .i.xg5!? ( l 0 .hxg5 with the idea i.e3.) 1 0 . . . �b6 1 1 . tLl b3 a6 1 2.h5 are promising for White.

A) S ... dS 9.i.bS J.d7 IO.adS �xdS In the ending that arises after

1 0 . . . tLlxd4 1 1 .i.xd7t �xd7 12.�xd4

tLlxd5 1 3 .tLlxd5 �xd5 14 .�xd5 exd5 1 5 .i.e3 i.e7 1 6.h5 White has slightly better prospects, Nikolenko - Zakharov, Smolensk 1 99 1 . 1 1 .tilxdS adS

Game 4S Karpov - Spassky Tilburg 1 980

12..te3!? This pawn sacrifice is more dangerous

for Black than 12 .�e2t. 12 .•. ie7

After l 2 . . . �xh4 13 .�d2!? ( l 3 .�e2 tLlxd4 14.i.xd4t �e7 1 5 .i.xd7t @xd7 1 6.i.e3 gd8 17.0-0-0 @c8 1 8 .�f3 a6=, Ashley - Salov, New York 1 996) 1 3 . . . a6 ( l 3 . . . i.e7 14 .0-0-0 with the idea tLl fS) 14 .i.xc6 bxc6 1 5 .0-0-0, White has sufficient compensation for the pawn. For example:

1 5 . . . i.d6 1 6. tLl fS i.xf5 17.gxfS i.e5 1 8 .f4! i.f6 19 .�b4! ( l 9 .i.c5 Nunn) .

2 12 The Scheveningen

Or 1 5 . . . ie7 Ferguson-Mirumian, European Youth Championship (U1 S) Chania 1 994. 16.if4!? (with the idea 1 6 . . . 0-0 17. liJf3! �f6 l S .g5 hxg5 19 .ixg5 �xf3 20.ixe7 E:feS 2 1 . 'lWg5 g6 22.�h6! E:xe7 23. E:h1 'lWf6 24.'lWh7t I!IfB 25. �hSt etc.) 16 . . . c5 17. liJ f3 ( 1 7. liJ f5!?) �f6 I S . E:de1 gives White a solid plus. 13.'ilYd2 ixh4

This is a risky idea. The exchanges with 1 3 . . . liJxd4

14 .ixd7t 'lWxd7 1 5 .�xd4 if6 1 6.�b4 are in White's favour, e.g. 16 . . . a5 ( 1 6 . . . ie7 17.'lWb3 with the idea 0-0-0) 17. 'lWa3 ie7 l S .�d3 'lWa4 1 9 .0-0-0! ? ( 1 9 . a3) 1 9 . . . 'lWxa2 20.'lWb5t I!IfB 2 1 . 'lWxb7 E:dS (Rowson - Stocek, Budapest 1 996) 22.�c7! E:eS 23.ic5 ixc5 24. 'lWxc5t I!IgS 25 .�a3;!; .

The best move for Black i s probably 1 3 . . . 0-0 when White has the following options:

a) 14 .ixc6 bxc6 1 5 .0-0-0 E:bS 1 6. liJ f5 ( 1 6.g5 might be better according to Shredder S. Unclear play arises after 1 6 . . . h5 17.g6 �b6 1 S . liJ b3 �a6 1 9 .id4 f6 20. l!Ib 1 .) ixf5?! ( 1 6 . . . if6 is better when Black is not worse after 17 .id4 E:eS.) 17.gxf5 if6 l s .id4 l!1h7 19 .'lWc3 ixd4 20. 'lWxd4, Nunn - ] . Anderson, Kilkenny 1 997, 20 . . . E:gS!? with equal play. The idea is to counter 2 1 . � xa7 with 2 1 . . . 'lWf6 (2 l . . . E:aS 22. 'lWxf7 E:xa2 is also very interesting) .

b) 14 .liJf5!? gives highly enterprising play: 14 . . . d4! ? 1 5 .ixh6 ib4! ( 1 5 . . . gxh6? loses to 16.'ilYxh6 if6 17.g5 ! geSt l S .@fl ihS 19 .96.) 1 6.c3 dxc3 17.bxc3 'De5! l S .ie2 geS 1 9 .@fl ifB 20.if4 'ilYa5 , Timmerman - Andersson, corr. 1 996, gave Black good counter chances.

c) White's most solid continuation is the relocation of the bishop to f3 with 14 .ie2. White doesn't have to fear 14 . . . ixh4 because he will win back the d5-pawn, and perhaps obtain slightly better chances. e.g. 1 5 .0-0-0 E:eS 1 6. liJ f5 ig5 17.ixg5 'lWxg5 l S .�xg5 hxg5 19 .if3 and White seems to be slightly better. 14.0-0-0 if6

The alternatives are dubious: 14 . . . liJxd4?! 1 5 .ixd7t 'lWxd7 1 6.ixd4

with the idea 1 6 . . . 0-0? 17.g5! hxg5 l s .ixg7! E:fcS ( 1 S . . . l!Ixg7 19 . E:xg5t) 19 .if6 'lWd6 20. E:xg5t ixg5 2 1 .'lWxg5t I!IfB 22. E:e1

14 . . . 0-0?! 1 5 .g5! liJxd4 ( 1 5 . . . ixg5 16 .ixg5 hxg5 17 . E:xg5 ! with the idea


CDxd4? 1 8 . �xg7t �xg7 1 9.'�hd4t f6 20. �gl t �f7 2 1 . WixdSt Jie6 22.Wixb7h-) 16 .Wi'xd4 hxgS ( 1 6 . . . JixbS? 17.gxh6! Jif6 1 8 . �xg7t �h8 19 .Wi'f4) 17.Jixd7Wi'xd7 18 .JixgS Jixgst 1 9 . �xgS f6 20. �xdS ± .

15.tLlf5 ixf5 16.gxf5 a6 17.,hc6t bxc6 IS.i.c5 gbS 19.M l::ib5?

Black decides to eliminate the bishop, but this sacrifice fails because of his inability to bring the h8-rook into play. Deserving of attention is 1 9 . . . �d7!? 20.c4 �c7 (20 . . . �c8? 2 1 .cxdS cxdS 22. Wie2 �b7 23. Wif3 �c6 24. �g4! with the idea �c4.) 2 1 .cxdS cxdS 22.Wid3 with an unclear position. 20.gge1 t cbd7 21 .c4 l::ixc5 22.bxc5 tgS

After 22 . . . Wi'b8 White wins by 23.cxdS JigS 24.�e3 Jixe3 2S .fxe3 Wi'eS 26.dxc6t �xc6 27.Wi'd7t �xcS 28.Wia7t �bS (28 . . . �c6 29.Wixa6t �c7 30.Wi'c4t �b8 3 1. Wi'b3t with the idea �d4+-) 29.Wi'b7t

�cS 30. �d2! Wial t 3 1 . �c2 Wi'xa2t 32. �dl Wi'a4t 33. �c2t �d6 34.Wic7t �dS 3S .Wi'cSt �e4 36. �e2. Therefore 22 . . . d4 was the best defence. 23.£4 Wi'f6

24.cxd5! Of course not 24.fxgS? Wial t 2S . �c2

Wi'xa2t 26. �d3 Wixc4t 27. �e3 hxgS . 24 ... Wi'al t 25.cbc2 YlYxa2t 26.cbd3 YlYxd2t?

The last mistake. Black could have offered more resistance by 26 . . . Wi'b3t 27.Wic3 WibSt 28.Wic4 Jih4 29. �hl Ji�. 27.l::ixd2 h£4 2S.l::ia2 adS 29.l::ixa6 h5 3o.cbd4 h4 31.cbxd5 l::ibS 32.f6 gxf6 33.gxf6 tg3 34.l::ixf7t cbdS 35JUSt 1-0

B) S • . . h5 9.gxh5 tLlxh5 10.tg5 tLlf6 The main alternative is 1 0 . . . Wib6 (The

passive 1 0 . . . Wi'c7 1 1 .Wi'd2 a6 1 2 .0-0-0 Jid7 13 .Jie2 CDxd4 14.Wi'xd4


i.c6 I S .i.g4! ? bS 1 6J'!:ge1 Vl!a7 17.VI!d3 ltJf6 ( 17 . . . VI!xf2? 1 8JW Vl!a7 19 .i.xe6! fxe6 20.eS) 1 8 .i.xf6 gxf6 1 9. 1tJ dS was very good for White in van Blitterswijk - van Beek, Haarlem 1 996.) l 1 . ltJ b3 a6 12 .i.e2 g6 (After 12 . . . ltJf6?! 13 .hS 'fffc7 14 .h6! Black has immediate difficulties as in the game Fedorowicz - S. Polgar, Wijk aan Zee 1 990, which continued 14 . . . ltJh7?! ( 14 . . . ltJd7) I S .hxg7 .txg7 1 6.i.e3 i.f6 17 .'fffd2 i.d7 18.0-0-0 i.e7 1 9J'!:hl 0-0-0 20. ltJ a4! with a large advantage for White.) 1 3 .'fffd2 and now:

a) 1 3 . . . VI!c7 14.0-0-0 bS I S .a3 i.d7?! ( 1 S . . . l'l:b8) 1 6.i.xbS!? axbS 17.lLIxbS Vl!b8 1 8 .lLIxd6t i.xd6 1 9.VI!xd6V1!xd6 20.l'l:xd6 l'l:c8 2 1 .l'l:gdl lLIb8 22.eS and the white pawns are stronger than the black knight, Govedarica - Mokry, Trnava 1 987.

b) 13 . . . i.d7 14.l'l:g2!? (White is not forced to sacrifice his pawn on £2) 14 . . . VI! c7 I S .0-0-0 bS 1 6.a3 lL1eS ( 1 6 . . . l'l:c8 17.f4!?) 17 .Wfd4 l'l:h7 1 8 .f4 lLIc4 1 9 .i.xhS l'l:xhS

20.l'l:e 1 ! i.c6 2 1 .lLIdS i.xdS 22.exdS eS 23 .VI!d3 i.e7 24. lLI d4! i.xgs 2S.hxgS 0-0-0 26.lLIc6 l'l:dh8, Motwani - Roca, Dubai (01) 1 986, and here 27.l'l:ge2!? l'l:h3 28.VI!dl gives White a clear plus, as 27 . . . exf4? does not work because of 28 .l'l:e8t l'l:xe8 29.l'l:xe8t �b7 30.VI!e4 l'l:xgS 3 1 . VI! d4+-.

Game 49 Morovic Fernandez - Veingold Spain 1 993

1 1 .�g3!? This move is less investigated. Usually

White has chosen 1 1 . 'fff d2 or 1 1 .i.e2 with the following possible continuations:

a) I 1 .Wfd2 Vl!b6 (after l 1 . . .a6 1 2.0-0-0 i.d7 1 3 .f4 Wfc7 14 .f5!? lLIxd4 I S .Wfxd4 i.e7 1 6.fxe6 fxe6 17 .i.h3 0-0-0 1 8 .Wfb4 (with the idea Wfb3) White's position is preferable, Santo Roman - Lepelletier, Cannes 1 994.) 1 2.lLIb3 a6 1 3 .0-0-0 i.d7 (Another


idea is 1 3 . . . Wfc7 14 .i>b1 b 5 when White can proceed 1 5 .8!? ib7 1 6.Wffl 0-0-0 17.ie3 tLld7 1 8 .l'k1 i>b8 1 9 .a4!? b4 20.tLla2 with the idea c2-c3 which will yield slightly better play, as in Frolov - Raisa, Helsinki 1 992. One point of White's play is that Black can't develop his dark squared bishop: 1 3 . . . ie7? 14 .ie3 and g7 hangs.) 1 4J�g3 (interesting is 14.i>b 1 intending to meet 14 . . . 0-O-O?! (14 . . . Wfc7) with 1 5 .ie3 Wfc7 1 6.f3 ! ie8 (16 . . . E1xh4? 17.Wffl) 17.Wffl tLld7 18 .ig5 tLlfG 1 9.f4! ie7 20.5±, Luther Ungureanu, Oberwart 2003.) 1 4 . . . Wfc7.

Now: a 1 ) in the game Karpov - Kasparov,

Moscow (3 1 ) 1 984, there occurred 1 5 . ig2 ie7 1 6.f4 0-0-0 1 7.Wffl c;t;>b8 1 8 .5 tLl e5 1 9 .ih3 tLl c4 20. tLl d2 tLl xd2 2 1 . E:xd2 E:c8 and Black had overcome his opening difficulties.

a2) 1 5 .Wfe2 b5 (in case of 1 5 . . . 0-0-o 1 6 .E:gd3 c;t;>b8 1 7.8 ie7 1 8 . Wffl

tLl e8 1 9 . c;t;> b 1 ixg5 20.hxg5 in Short - Xu Jun, Lucerne 1 985 , White could claim a slight advantage) 1 6 .a3 E:c8 17 .c;t;>b 1 b4! ? (stronger than 1 7 . . . tLl e5 ?! 1 8 . f4 tLl c4 1 9 . E:gd3 tLl h7 20.5! with initiative, Kindermann - Briffel, Dubai (01) 1 986) 1 8 .axb4 tLl xb4 Gallagher - Pritchett, BCF (ch) , Swansea 1 987, and now 1 9 .ixfG gxfG 20.f4 reaches a position which is more difficult to handle for Black because his king is permanently stuck in the centre.

a3) 1 5 . ie2 (threatens h4-h5-h6) 1 5 . . . b5 ( 1 5 . . . ie7? ! 1 6.if4 g6 1 7.ig5 and White is better, Ziegler - R. A kesson, Swedish Championship 1 999 .) 1 6.a3 b4 1 7.axb4 tLl xb4 1 8 .h5 Black's position is critical.

1 8 . . . d5 ( 1 8 . . . tLl xh5 1 9 . E:h 1 tLl xg3 20. E:xh8 tLl xe2t 2 1 .Wfxe2 tLl c6! (2 1 . . . fG 22.ixfG, White wins as in Ziegler - Maus, Lugano 1 988 . ) 22.f4 tLl e7 23 .ixe7 c;t;>xe7 24.5 White is slightly

2 1 6 Th e Scheveningen

better.) 1 9 .h6 dxe4 20. lt>b l a5 2 1 . tLl b5 �dB 22. tLl d6t It>e7 23. tLl xe4 tLl bd5 24.c4 tLl b6 25 . tLl xf6 gxf6 26.�xf6t 1-0 Ziegler - Ludvigsen, Gausdal 1 9BB .

b ) 1 1 . '!e2 a6 12 .�d2 (White has nothing after 1 2.h5 .!d7 ( 1 2 . . . .!e7!?) 1 3 .�d2 .!e7 14.0-0-0 tLlxh5 1 5 . .!xh5 E1xh5 16 . .!xe7 �xe7 1 7. E1xg7 0-0-0) 1 2 . . . �b6 (Black can try 1 2 . . . .!d7 1 3 .0-0-0 b5 14 . tLl xc6 .!xc6 1 5 . �e3 �a5 1 6 .lt>b l .!e7. However, White is clearly better after 1 7.e5 ! . ) 1 3 . tLl b3 .!d7 1 4.h5 ! ? ( 1 4 . E1g3!, which can transpose to the variation above, seems to be best. Actually, the idea 14 . E1g2! , which was used in Motwani - Roca above, is also interesting.) 1 4 . . . tLl xh5! 1 5 . E1h l g6 1 6.0-0-0 �xf2! 1 7.e5 ! ? and now, instead of 1 7 . . . �f5! ? with unclear complications, an interesting idea is 1 7 . . . tLl xe5 ! ? I B . tLl e4 �f5 1 9 .�e3 ( 1 9 . E1xh5? gxh5 20. tLl f6 t It> dB 2 1 . tLl d5 t It>cB 22. tLl b6t It>bB-+) 1 9 . . . �c6 20. tLl f6t (after 20. tLl bd2?! tLl d7! White does not have enough compensation for the sacrificed material) 20 . . . tLl xf6 2 1 . E1xhB tLl eg4 22 . .!xg4 tLl xg4 23 . � d2 .!d5 with better prospects for Black. 1 l . . . a6

In this situation the standard reply 1 1 . . . �b6 is dubious in view of 1 2. tLl db5 ! tLl e5 ( if 1 2 . . . a6? , then 1 3 . tLl xd6t .!xd6 14 .�xd6 �xb2 1 5 . E1b l �xc2 1 6.'!xf6 gxf6 1 7 . .!c4 with the idea .!b3) 1 3 .f4! (less clear is 13 . .!e3 �dB 14 . tLl xa7 E1xh4) 1 3 . . . tLl g6 1 4.e5! dxe5 1 5 .fxe5

tLl h5 ? ( 1 5 . . . tLlxe5 16 .�e2 tLl fd7 1 7.0-0-0f) 1 6. E1d3 .!e7, Hellers - van WeIy, Wijk aan Zee 1 990. Now White has 1 7. tLl d6t! It>ffi I B . .!xe7t It>xe7 1 9 . tLl f5t! It>eB 20. E1d6 winning.

And after 1 1 . . . .!e7 1 2�d2 tLl xd4 1 3 . �xd4 �b6?! 1 4 . .!b5t! It>ffi 1 5 .�xb6 axb6 White can exploit the weak b-pawns by 1 6 .0-0-0 e5 17 .E1gd3! .!e6 I B .8, Gavrikov - Adorjan, BieI 1 990, planning b3 and tLl a4. 12.tL\xc6!? bxc6 13.YlYa

White prepares queenside castling as quickly as possible, when Black has to reckon with the advance e4-e5 .

13 . . J!�b8 This is probably the best. The

alternatives are: a) 1 3 . . . e5? ! 1 4 . .!c4! .!e6 1 5 . .!xe6

fxe6 1 6.0-0-0 E1bB 1 7.h5! E1b7? ( 1 7 . . . �b6?! I B .b3 tLlxh5? loses to 1 9 . E1h l g6 20 . .!e3 followed by E1xg6. Black should have tried 1 7 . . . .!e7 after


which White can continue 1 8 .h6!? gxh6 1 9 .i.h4 �ffi 20. �g7 'lWb6 2 1 .b3 with the idea 2 1 . . . �b7 (2 1 . . . �fl 22. �xfl Wxf'7 23. �xd6!) 22.'lWh3.) 1 8 .h6 �fl 1 9 .hxg7 i.xg7 20.'lWd3 d5 2 1 .'lWxa6 'lWc7 22. lLl b5 ! cxb5 (Or 22 . . . 'lWd7 23.'lWa8t We7 24. 'lWa3t Wd8 25 . lLl d6 �e7 26. �b3 White wins.) 23 .�c3 , 1 -0 Vasiukov - Danailov, Moscow 1 986.

b) 13 . . . i.d7?! 14 .0-0-0 i.e7 1 5 .e5! dxe5 1 6. lLl e4 with compensation for the pawn, e.g. 1 8 . . . �b8 17 .i.xf6 gxf6 ( 1 7 . . . i.xf6? 1 8 . �xg7!) 1 8 .'lWg2! �ffi 19 . .tc4 (with the idea 19 . . . �b4 20.i.xe6! fxe6 2 1 . �g8 'lWc8 22. �xffit Wxf8 23 . �gl We8 24. 'lWg8t i.ffi 25 . �g7 with a decisive advantage for White) 1 9 . . . 'lWb6 20. �b3 'Wa7 2 1 . �bd3 �b7 22.'lWg7 and White's initiative proved decisive in Kir. Georgiev - Sax, Reggio Emilia 1 988/89.

c) 13 . . . i.e7 14 .0-0-0 d5 (after 14 . . . 'lWc7 - 1 5 .e5 ! dxe5 1 6.h5!? (with the idea h6) 1 6 . . . lLl xh5 17 .i.xe7 Wxe7 1 8 . �h3 g6 1 9 . 1Ll e4 5 20.'lWa3t Wfl 2 1 . lLl d6t Wf6 22.i.e2 gives White a strong attack, Sibarevic - Masic, Pula 1 990) 1 5 . .tc4!? .tb7 1 6 . .tb3 Wffi 1 7.h5!? lLl d7 ( 17 . . . lLlxh5? ! would be met by 1 8 .i.xe7t Wxe7 ( 1 8 . . . 'lWxe7? 1 9 . �hl g6 20. �xg6) 1 9 .exd5 cxd5 20 . .txd5! exd5 2 1 . lLl xd5t .txd5 22. �xd5 lLl xg3 23. �xd8 �axd8 24.fxg3 with good winning chances for White) 1 8 .i.f4t , Pavlovic - Razuvaev, Cetinje 1 99 1 . 14.0-0-0 �b6 15.b3 lLlh5

Less accurate is 1 5 . . . i.e7 because of

1 6.e5!? dxe5 1 7. lLl e4 i.a3t ( 1 7 . . . lLl h5? 1 8 .i.xe7 Wxe7 1 9 . �xg7! and 1 7 . . . lLlxe4 1 8 .'lWxe4 White's compensation is ample. ) 1 8 . W b l lLl d5 1 9 .c4 i.b7 20.cxd5 cxd5 (with the idea 2 1 . lLl d2? 'lWd4 22. lLl c4 dxc4. ) . This was played in Gorin - Rodin, Simferopol 1 989. Now strong was 2 1 . 'lWf6! dxe4 22.'lWxg7 �ffi 23.'lWxe5 �c8 24. �c3 ± with the idea 24 . . . �xc3? 25 .i.b5t! axb5 (25 . . . i.c6 26.'lWxc3) 26.'lWxb5t+-. 16J:'!gl

16".d5 White did not achieve anything in

the game van der Wiel - Winants, Budel (zt) 1 987, after 1 6 . . . g6 17 .i.c4?! 'Wa5 1 8 .e5?! d5, but instead of 17.i.c4?! both 1 7. i.h3 and 17. W b 1 deserved attention. The advance d6-d5 looks rather risky because White's forces are better prepared for concrete play and the position of the black king in the centre becomes more vulnerable.

2 1 B The Scheveningen

17.<1>b1 ib4?! Also if 17 . . . id6?! then I B .ih3 causes

serious problems, e.g. I B .. JMc7 ( 1 B . . . ie5 1 9 .exd5 cxd5 20. ggel with the idea 20 . . . f6 2 1 . ig4 g6 22.ixh5 gxh5 23. liJxd5 exd5 24.�xf6) 1 9 .exd5 cxd5 20. gge 1 ie5 (20 . . . ib7 2 1 . ixe6 fxe6 22. gxe6t c;t>d7 23.1lMf5) 2 1 . liJxd5 exd5 22.1lMxd5 ixh3 (22 . . . f6 23. gxe5t!) 23. gxe5t c;t>f8 24. ge7+-.

Black should have sought to prevent the eventual sacrifice on d5 by 17 . . . ib7, although White's position is more promising after, for example I B .ih3 . IS.ih3! �a5?

Missing a second chance to play ib7.

19.ti:)e2!? A narural reply, although 1 9.exd5 ! ?

was strong too as White wins a pawn after 1 9 . . . cxd5 ( 1 9 . . . ixc3? 20.dxe6+-) 20. liJxd5! i.b7 (20 . . . exd5? 2 1 . gxd5 1lMc7 22.ixcB gxcB 23.1lMe4t) 2 1 .if4! ixd5 (2 1 . . . gcB 22. liJf6t liJxf6 23.1lMxb7 �c5 24.c4) 22. gxd5 �xd5 23.�xd5 exd5 24.ixbB± . 19 ••• ,td7?

Black is already in trouble and it is hard to suggest anything else. 20.c4

Another, and perhaps better possibility, was 20. liJ f4 liJxf4 2 1 .ixf4 gb5 22. gxg7. 20 ••• dxe4 21.�xe4 'ilYc7?!

More stubborn is 2 1 . . . liJ f6 22. � f4 gb7. 22.ti:)f4! ti:)f6 23.'ilYe2

23.ixf6 gxf6 24. gg7! (Threatening i.xe6) would have won easily with the idea 24 . . . gxh4 (24 . . . �e5 25 .�d3) 25 . ggBt if8 26. liJ g6!. 23 •• .id6?

The final mistake. 23 . . . gdB was the only way to continue the fight. 24 • .he6! ixe6 25.ti:)xe6 fxe6 26.'ilYxe6t ie7 27.ixf6 gxf6 28.ggSt gxgS 29.'ilYxgSt 1-0


The Kalashnikov

- By Jan Pimki

1 .e4 c5 2.�f3 �c6 3.d4 cxd4 4.�xd4 e5 5.�b5 d6

The Sicilian Kalashnikov has become a real opening over the last 1 5 years or so. It was developed by Evgeny Sveshnikov more than anyone to start with, but he already had his own main line in the Sicilian!

Here I suggest 6.� lc3 is the right decision. 6.c4 is the main theoretical move, but it promises positional play with an easy way for Black to gain equality. Instead this more tactical and confrontational approach promises White an advantage, and makes me wonder how I ever got away with playing the Kalashnikov for so long without being punished!

But as this is a minor line, I will not waste your time with unnecessary talk. All you need to know is in the games below. Basically you play 8 .lLld5 , 9.exd5 and 1 0 .c4!, or the same moves in another order!

Game 50 Motylev - Shariyazdanov Tomsk 2004

1.e4 c5 2.�f3 �c6 3.d4 cxd4 4.�xd4 e5 5.�b5 d6 6.� 1c3 a6 7.�a3 b5

The alternatives are: 7 . . . i.e6 8 . lLl c4 b5 9 . lLl e3 lLlf6 1 0.g3!

The best way for White to play for an advantage. 1 0 . . J'l:c8 1 1 .i.g2 and White is slightly better.

7 . . . i.e7 8 . lLl c4 b5 9 . lLl e3 lLl f6 10 .g3! h5! ? GM Teimour Radjabov's idea. ( 1 0 . . . 0-0 l 1 .ig2 ie6 12 .0-0 White is slightly better.) 1 1 .i.g2 h4 1 2. lLl ed5! A new move. 12 . . . lLlxd5 1 3 .lLlxd5 h3 14 .iB lLld4 1 5 .ie3 lLlxf3t 16.�xf3 E:b8 1 7.0-0 i.b7 1 8 . E:adl White is a little better. 8.�d5 �ge7

The most popular move here. There was a time when I believed in this position for Black, but I have changed my mind. White is better but the position is extremely complicated.

8 . . . i.e7 9.c4 b4 10. lLl c2 ( 1 0 .lLlxb4? lLlxb4 1 1 .�a4t i.d7 12 .�xb4 d5 13 .�c3 �b8 and Black is better.) 1 0 . . . E:b8 l 1 .b3 lLlf6 12 .i.d3 0-0 1 3 .0-0 lLld7 14 .i.b2

220 The Kalashnikov

ttJ c5 1 5 . ttJ ce3 i.g5 16 .i.c2 a5 17 .r;t>h1 White i s better.

8 .. .:1'J:b8 9.c4 ttJge7? (9 . . . b4 - see 8 . . . i.e7) . For nearly ten years this position has been assessed as unclear in ECO. The game Kaminski - Pinski, Katowice 1 995, and the winner's comments (GM Kaminski) were the basis for this. The truth is different: 1 0.cxb5 ttJxd5 I l .exd5! ttJd4 1 2.bxa6 'lMfa5t 13 .i.d2 �xd5 14.a7 �a8 1 5 .i.e3 �xa7 1 6. ttJ b5 with a clear advantage for White.

9.c4 It is easy to remember. No matter what

Black play (8 . . . i.e7, 8 . . . ttJge7, 8 . . . ttJce7, 8 . . . �b8), White should always answer c2-c4. 9 ... �xd5

The more popular move order is: 9 . . . ttJd4 1 0 .cxb5 ttJxd5 I l .exd5. 10.exd5

10 . �xd5 ! ? i.d7 1 1 .i.e3! White is at least slightly better. I l .cxb5? ! ttJd4!

12 .i.e3! i.e6 13 .�b7 �b8 14 .�xa6 �a8 1 5 .�b7 �b8 is equal. 10 ... �d4 11 .cxb5! J.d7

The only move, after which we reach the critical position for the 8 . . . ttJge7 line. 1 1 . . . �h4? There was a time when people believed in this move. 1 2.bxa6!! I was unlucky enough to face this innovation. (Previously the weak 1 2.i.e3? had been played.) 1 2 . . . i.e7 13 .i.e3 0-0 14 .i.xd4 exd4 1 5 .i.d3 �xa6 16.0-0 and White has a winning position, Iordachescu-Pinski, Cappelle la Grande 1 998. 1 1 . . .i.e7?! 1 2.i.c4! 0-0 ( 1 2 . . . 'lMfa5t 1 3 .i.d2 'lMfb6 14.bxa6! 'lMfxb2 1 5 . �b1 �xa3 16.i.b4 i.g4 17.i.xa3 i.xdl 1 8 . r;t>xdl and White wins.) 1 3 .0-0 f5 14.bxa6 f4 1 5 . ttJ c2 ttJxc2 16 .�xc2 f3 17.i.d3! White has a clear advantage. 12.J.e3

Just weak is 1 2.bxa6? 'lMfa5t 1 3 .i.d2 'lMfxd5. 12 ... axb5

Black can't be happy after 12 . . . ttJxb5 1 3 . ttJ c4! Remember this useful idea: when the black knight takes on b5, the white knight comes to c4. 1 3 . . . f5 14.a4 ttJa7 1 5 . �c1 f4 16.i.d2 ttJ c8 17.�b3 and White is better, Smirin-Shirov, Klaipeda 1 988. 1 2 . . . i.e7 13 .i.d3 axb5 ( 1 3 . . . ttJxb5 14. ttJ c4! etc.) 14 .0-0 �b8 1 5 .i.xd4 exd4 1 6 . ttJ c2 0-0 17 .ttJxd4 'lMfb6 Here I have a new idea: 1 8 .i.f5! (Instead of 1 8 . �c1 ) 18 . . . i.f6 19 .i.xd7 i.xd4 20.�d2 and White is better. 13 . .ixd4 exd4 14.J.d3 Wa5t 15.<!>f1

The posltlon after I S . �d2 !i.e7 1 6.�xaS E1xaS 17. t2k2 is not so clear. Black has some compensation for his pawn weaknesses here. 15 .. ..te7

I S . . . b4 1 6. li:l c4 �cS 17 .�e2t i>d8 1 8 .�f3 !i.bS 1 9 .b3 E1a7 20.g3! hS 2 1 .h4 and White is better. 16.'i!Ye2 E1b8 17.tLle2 <it>f8 18.tLlxd4 'i!Yb6

19.tLlOO! Opposite coloured bishops do not

always lead to an easy draw, they can also give good attacking possibilities. 19 ..• ixOO 20.dxOO 'i!YxOO 21 .h4 h5 22J�eI if6 23.gh3 gh6

This looks ineffective, but it is hard to find alternatives. 23 . . . g6? 24.!i.xg6! fxg6 2S .E1f3 i>g7 (2S . . . dS?? 26. �e7t and White wins.) 26Jhf6 E1he8 27. E1e6 E1xe6 28.�xe6 �c4t 29. i>gI This endgame is winning for White. 24.ge3

White has an interesting plan: E1e4-b4-bS . 24 .•• ie5

24 . . . b4?? 2S .!i.bS and White wins. 24 . . . g6 2S . E1e4 dS 26. E1b4 �d6 27.�d2 i>g7 28. i>gl E1hh8 29.!i.xbS !i.xh4 30. E1xh4 E1xbS 3 1 . E1d4 with good winning chances. 25.ge4 g6

There is no time for 2S . . . b4 26.f4! E1f6 27. �xhS and White gains a decisive advantage. 26.gb4 <it>g7 27.gxb5 ghh8 28.b3 gxb5 29.ixb5

In positions like this White can continue to play for a win forever, without having to take any risks. 29 ••• 'i!Ye5 30..te4 'i!Ya3 31 .g3 gaS 32.'i!Ya ga7 33.ge2 .ib2 34.'i!Ye3 1f6

34 . . . �xa2?? 3S . E1xb2 and White wins. 35.'i!Yd2 ga5 36.<it>g2 'i!Ye5 37.<it>gl 'i!Yb6 38.ge3 ga3 39.ga ga7 40.a4 'i!Ye5 41 .gd3 1e5 42.gd5 'i!Ya3 43.<it>g2 'i!Yal 44.'i!Ydl 'i!Ya2?

The position was more or less lost, but after this mistake all is clear. 45.gxd6! 'i!Yal 46.'i!Yxal inl 47.gd3 ib2 48.<it>a .ia3 49.�e4 .ie5 50.£4 ib4 5 U�d5 .lei 52. <it> a 1e3 53.£5 1b4 54..td3 ge7 55.gb5 .ie5 56.b4 ie7 57.fxg6 fxg6 58.gb6 gc3 59.<it>e2 gb3 60Jixg6t <it>f'7 61 .ga6 gb2t 62.<i)fl ixb4 63.ig6t <it>e7 64.gb6 1-0

222 The Kalashnikov

Game 51 Delchev - Brumen Zaclar 2003

1 .e4 c5 2.�f3 �c6 3.d4 cx:d4 4.�xd4 e5 5.�b5 d6 6.� 1c3 a6 7.�a3 b5 8.�d5 �ce7!? 9.c4! �xd5 10.ad5!

Keeping the symmetrical pawn structure is less dangerous for Black. White has some problems with the knight on a3 after cxd5. 10 . . . bxc4

There is no compensation after 10 . . . !e7 1 1 .cxb5 �f6 12.ic4 0-0 13 .bxa6 ixa6 14.0-0 �b6 1 5 .�d3 E:fcS 1 6.b3 h6 17. E:d 1 White is clearly better. 1 1 .�xc4 �f6

1 l . . .f5 12 .,td2 E:bS ( l2 . . . a5 ? 13 .�a4t !d7 14 .�a3 �bS 1 5 .!xa5 White take a pawn.) 13 .ia5 �e7 14. ttJb6 !b7 1 5 .�a4t 'it>f7 16.g4! fxg4 17.h3 White has a clear advantage. 12.J.e3!

This is more dangerous than 1 2.!d2. 12 . . ':!�b8 13.J.e2 ie7 14.0-0 0-0 15.a4

The critical position for the S . . . ttJce7 variation. 1 5 . . . !b7 seems to be the best way to search for equality. 15 ..• ib7

1 5 . . . ttJd7 1 6.a5! This is better than the previously popular 1 6.Qd2. White's chances are on the queenside, therefore it makes no sense to play slowly. 1 6 . . . f5 17.f4 g5 ( l7 . . . exf4 1 S .ixf4 ttJc5 19 .ttJb6 !b7 20.b4 ttJ e4 2 1 . E:cl and White is better.) l S .fxe5 ttJxe5 1 9 .ttJb6 ,tf6 20. ttJxcS �xcS 2 1 . E:cl �d7 22.!b6 White is slightly better, Ehrenfeucht -Pinski, Warsaw 1 999.

15 . . . �d7 16. ttJb6 �f5 17 .ttJxcS �xcS l S . E:cl �f5 19 .b3 a5 20.h3 e4 2 1 .!c4 and White is better.

1 5 . . . a5 ?! 1 6.!d2 E:aS 17 .�e1 with a clear advantage. 16.�b6 �d7

16 . . . �eS 17.a5 !dS l S .!c4 !xb6 1 9 .axb6! ( l9 .!xb6 ttJd7 20.!e3 f5 with counterplay.) 1 9 . . . ttJd7 20.�b3 f5 2 1 .f3 f4 22.!f2 White is a bit better - he has the two bishops and an attack on the a6 pawn, Bogachkov-Sherbakov, Russia 200 1 . 17.a5 f5 18.f3 We8

l S . . . ttJxb6 19 .!xb6 �d7 20.b4 !dS 2 1 .,te3 if6 22. E:b1 �f7 23.ic4 E:bcS 24.�d3 E:aS 25 .E:fcl �h5 26.ib3 ig5 27.�d2 !xe3t 2S. �xe3 White is slightly better, Karjakin - Kosteniuk, Brissago (4) 2003. 19.Wd2

More thematic is 1 9 .b4 .tdB 20. ttlxd7 'l&xd7 2 1 . l"!:b l and White has the more pleasant game. 19 ••. .idS 20.�xd7 'ilYxd7 21 .b4 f4?

If such a move does not win by force then it is a bad move! But seriously: Black breaks his own pawn chain, irreversibly, and without any compensation! 22 . .ifl 'ilYf7 23.lUdi J.g5 24.J.d3 'ilYh5 25.b5 ixd5?!

25 . . . axb5 26.a6 .taB 27 . .ta7 l"!:bdB 2B . .txb5 White is simply better.

26.ixh7t! �xh7 27.'ilYc2t! 27.'I&xd5? l"!:xb5 2B.'I&e4t 'l&g6

29.'I&xg6t <;!;>xg6 30. l"!:xd6t l"!:f6 and Black is alive! 27 .•• �hS 2SJ:�xd5 axb5 29J'lxd6

Now White has good attacking possibilities because of the weak light squares. White also has good possibilities to win by promoting the a-pawn. 29 • . • J.e7 30J�b6 b4 31.gxbS gxbS 32.a6

Brutal and effective. Black does not have any chance. 32 .• .i.h4 33.J.b6 .idS 34.ixdS gxdS 35.a7 'ilYe8 36.'ilYe4 gaS 37.'ilYxb4 'ilYd7 3S.YlYbSt �h7 39.'ilYb6 'ilYd5 40.h3 e4 4I J!a5 'ilYf7 42.YlYbS 1-0

Game 52 Anand - Shirov Linares 2002

The annotations to this game are inspired by those of Anand, as published in several places. 1 .e4 c5 2.�f3 �c6 3.d4 cxd4 4.�xd4 e5 5.�b5 d6 6.� 1c3 a6 7.�a3 b5 S.�d5 �f6 9.c4

This is the most aggressive line. 9 . .tg5 transposes to very well known positions from the Sveshnikov Sicilian, and not the lines recommended in this book. 9 ... b4

9 . . . ttl d4 leaves us with some options:

224 The Kalashnikov

a) 1 0.cxb5!? ltJxd5 l 1 .exd5 transposes to the lines after 8 . . . ltJge7 9.c4 ltJd4 1 0.cxb5! ltJxd5 l 1 .exd5 .

b) If you do not like complications, try 10 . .id3!? ltJxd5 l 1 .cxd5 g6! (This looks better than 1 1 . . . .ie7 12.0-0 0-0 1 3 . .ie3 This is a common position for the 8 . . . ltJ f6 and 8 . . . ltJge7 variations. (If Black takes ltJxd5 of course) White has the advantage. 1 3 . . . .id7 14.Wld2 Wlb8 1 5 J �ae 1 .id8 16 . .ixd4! exd4 17. ltJ c2 .if6 1 8 .f4 etc.)

12. ltJ c2 (Interesting is 12 . .ie3 .ig7 1 3 .0-0 0-0 14 .ixd4 exd4 1 5 . ltJ c2 with the plan ltJ b4-c6. However White should play very carefully because Black has good attacking chances here. 1 5 . . . Wlh4! 1 6.a4! ( 1 6. ltJ b4?! f5 17.exf5 ixf5 1 8 .a4 .ig4! 1 9 . .ie2 ( 19 .Wlc2? .if3!! 20J Hei (20.gxf3? ie5 2U�fe1 Wlxh2t 22 . 'it>f1 Wlh3t 23. 'it>e2 Wlxf3t Black is winning.) 20 . . . Wlg5 (20 . . . .ixg2 2 1 . 'it>xg2 Wlg4t 22. 'it>f1 Wlh3t 23 . 'it>gl Wlg4t Draw.)

2 1 .g3 gac8 with initiative.) 1 9 . . . .ixe2 20. Wlxe2 with unclear play.) 1 6 . . . bxa4 17. gxa4 .id7 1 8 . ga5 f5 1 9 .exf5 .ixf5 20.b4 White is at least slightly better.) 12 . . . ltJxc2t 1 3 . .ixc2! (After 13 . Wlxc2 .id7 the black bishop keeps control of the a4 square.) 1 3 . . . ig7 14.a4 id7 1 5 .ie3 0-0 16.0-0 f5 17.f3 and White is slightly better because of his space advantage.

c) l O . .ig5! Wla5t (Of course there is no real compensation after 1 0 . . . ie7? 1 1 ..ixf6 .ixf6 12.cxb5) 1 1 ..id2 Wld8 12 .cxb5 ( 12 .ltJxf6t Wlxf6 13 .ic3 ib7 leads to an unclear position.) 12 . . . ltJxd5 1 3 .exd5 Wlb6 14.ie3! ( 14. ltJ c4 Wlc5 1 5 .bxa6 .ixa6 1 6. ltJ e3 .ixfl Black has good compensation for the pawn.) 14 . . . .ie7 1 5 . .ixd4 exd4 16 . .id3 .ib7 17. WI f3! White has a clear advantage. For example: 1 7 . . . WI c5 1 8 .0-0 ixd5 19 .Wlg4 0-0 20. gae l ! gfe8 2 1 .b4! Nice trick. 2 1 . . .Wlxb4 22.Wlf5 g6 23.Wlxd5 and White wins.

9 . . . ltJxe4?! 1 0.cxb5! Wla5t 1 1 ..id2 ltJxdl 12 . Wlxd2 axb5 13 . ltJxb5 Wlxd2t 14 . 'it>xd2 'it>d8 1 5 . .ic4 White has the advantage. lO.tLlxf6t

Of course not 10 .ltJxb4?? because 1 0 . . . Wla5 l O. ltJ c2!? ltJxe4 ( 1 0 . . . gb8 This is better than l O . . . a5 , after which the b6 square is weak. l 1 .id3 ie7 12 .b3 White stands slightly better - compare the line 8 . . . .ie7 9.c4.) 1 1 . WI f3 ltJ c5 ( 1 1 . . . f5 12.id3 ile7 13 .ilxe4 fxe4 14 .Wlxe4 gb8 1 5 . Wlf3! White is better.) 1 2 .ltJdxb4


lDxb4 1 3 .lDxb4 !b7 14. lDd5 !e7 1 5 .b4 lDe6 1 6.!e3 0-0 17.!d3 !xd5 l S .cxd5 White is better, Mitkov - Permuy Lorenzo, Nigran 1 994. 10 • • .'IWxf6 1 1 .�c2 YlYg6

This position has been played a few times. The conclusion seems to be that White has the better chances. White has two possibilities now: 12 . 'IW d5 leads to complicated play; 1 2. lD e3 is less ambitious, but White is safer here.

12.YlYd5! The winner of the following game

wrote in New In Chess that he found no faults in Anand's play, but all the same he decided to play more carefully and secure a simpler advantage with the following approach: 1 2. lD e3!? !e7 ( 12 . . . !b7 13 .!d3 !e7 14.0-0 0-0 1 5 . lD f5 E1fdS 1 6.!c2 'lWe6 17 . .te3 g6 l S .lDxe7t lDxe7 1 9 .b3 f5 20.!g5 f4 2 1 .a3 bxa3 22. E1xa3 and White is much better, Palac - Srebrnic, Ljubjana 2004.

12 . . . 'lWxe4 13 .!d3 'lWd4 14 . lDd5 E1bS 1 5 .0-0 'lWg4 16 .'�a4 'lWd7 17.!e3 !e7 l s .lDb6 'lWc7 1 9 .1Dxcs E1xcS 20.'lWxa6 and White is slightly better, according to grandmaster Viorel Bologan.) 1 3.g3! 0-0 ( 1 3 . . . h5? 14 .!g2 h4 1 5 . lD f5 !e6 16 .!e3 E1cS 17 .lDxe7 �xe7 1S .c5 White is much better according to grandmaster Alexei Fedorov.) 1 4 . .tg2 !g5 1 5 .0-0 !xe3 ( 1 5 . . . lD d4?? 1 6.f4 !h6 17.f5 'lWg5 l s . lD d5 'lWdS 1 9 .!xh6 gxh6 20.f6 �hS 2 1 . 'lWh5 and White has a decisive advantage.) 1 6.!xe3 E1bS 17 . 'lWa4 !b7 l S . E1fdl a5 19 . E1d2! White stands better, Bologan-Tregubov, Belfort 2002. 12 ••• .ib7 13.�xb4 gc8

1 3 . . . lD dS 14 .'lWdl ! 'lWxe4t 1 5 .!e3 !e7 16 .lDd5 E1cS ( 1 6 . . . .th4 17 .'lWd3 .txd5 ( 1 7 . . . 'IW xd3 I S . !xd3 is also better for White.) l S .cxd5 'lWb4t 19 .'lWd2 'lWe4 20. E1cl .tg5 2 1 . 'lWd3 'lWb4t 22.'lWc3 'lWxc3t 23. E1xc3 !xe3 24.fx:e3 This is one of the typical endings in the Kalashnikov. It is also much better for White. First of all, White has bishop against knight. Secondly, the black pawn on a6 is weak. Last but not least, White has good chances to create a passed pawn on the queenside.) 1 7.£3 'lWg6 l S .lDxe7! (Anand believes that after I S . 'IW d2 White is much better. However it is not so easy to prove. l S . . . !xd5! 19 .'lWxd5 (Of course bad is 1 9.cxd5 E1c2 20.'lWb4 0-0 and White is even worse.) 1 9 . . . lD c6 and White has serious weaknesses on the kingside. The question is: how does White finish his

226 The Kalashnikov

development? For example: 20.�d2? d5! with an attack.) 18 . . . @xe7 19 .�d2 and White is much better! 14.,ie3

14.f3 ie7 1 5 . lDxc6 ixc6 1 6.�d2 f5 17.exf5 �xf5 1 8 .ie2 0-0 1 9 .0-0 e4 with compensation, according to Anand. 14 ... i.e7

14 . . . lDd8 1 5 .�d2 �xe4 1 6.f3 �h4t 17.if2 �h5 1 8 .lDd5 �g6 1 9 .1'!d1 ixd5 20.cxd5 ie7 2 1 .ig3 White stands better. 15.h4!

A strong prophylactic move, which is also a strong attacking move. Prophylactic, because White takes control of the g5 square. Attacking, because White threatens h4-h5. 15 ... 0-0

1 5 . . . �g4 Black defends 1'!c8 and threatens lDxb4. 16 .1'!d1 0-0 17.f3! ixh4t 1 8 .@e2! �g6 19 .1Dxc6 ixc6 20. �xd6 if6 2 1 . �a3 White's position is

much better: there is no real compensation for the material.

1 5 . . . h5? ! 1 6.f3 0-0 17.0-0-0 lDxb4 1 8 . �xb7 lDxalt 19 . @c2 White stands much better - Anand. 16.h5 YlYg4 17.a YlYg3t

17 . . .'�'d7 1 8 .lDxc6 ixc6 1 9 .�d2 White is better.

18.�dl 1 8 .if2? lDxb4 19 .�d2 �g5 20.�xb4

d5! 2 1 .c5 The only move. 2 l . . . 1'!b8 22.exd5 ixd5 with a strong initiative. 18 . . .i.g5 19.tthc6

19 . 1'!h3 �xh3 20.gxh3 ixe3 2 1 . lD c2! (Better than 2 1 . �d3 id4 22. lD d5 h6! with positional compensation for the material, according to Anand.) 2 1 . . .ib6 (2 1 . . .if4 22.b4 1'!b8 23.�d3 f5 24.a4 lD e7 25 .a5 White is better.) 22.h6 gxh6 23.�d2 and White is better. 19 ... ixe3

1 9 . . . ixc6 20.�d2 ixe3 2 1 . �xe3 f5 22. 1'!h3 �f4 23.�xf4 exf4 24.exf5 1'!xf5

25.b4! White has a clear advantage: Black has a wrecked pawn structure, and not enough active play to compensate for it. It is only an illusion that the rook on h3 is out of the game. After �h4 (xf4) it is playing at 1 00%! 20.tOe7t <bh8 21 .tlYd3

2 1 .�xb7? 'lMff2 22 . .ie2 (22.�d5 �xb2 Black has a clear advantage.) 22 . . . �b8 23. 'IMf d5 �xb2 24. �e 1 .ig5 Black is close to winning. 21. • .J.g5 22.tOxc8

22. lt:lf5? This does not prevent Black's counterplay. 22 . . . �f2 23.�c2 'lMfb6 with a strong initiative for the pawn. 22 . • . gxc8

23.h6?! 23. <t!ic2 �f2t 24 . .ie2 d5 25Jl:afl !!

(25 .exd5? ! .ixd5 26.b3 �e6 27. �af1 �c5 28. �d1 e4 29.fxe4 'lMfa3 30. <t!ib1 .ixc4 3 1 . 'lMfc2 .ie6 32.'lMfb2 �a5 33Jl:hfl h6 with good practical compensation.) 25 . . . dxc4 (25 . . . 'lMfc5 26.exd5 �xd5 27. <t!ib1

and White wins.) 26.'lMfd8t! �xd8 27. �xf2 .ie3 (27 . . . �d2t 28. <t!ic3 .ie3 29.g3 and White wins.) 28. �hfl �d4 29.g3 .ixf2 30. �xf2 �c8 3 1 . �fl .ie6 32. <t!ic3 f6 33. �d1 Eixd1 34 . .ixd1 This ending is just lost. 23 ••• gxh6 24.<bc2 dS?

24 . . . f5 25.exf5 'lMff2t 26 . .ie2 d5 27. Eiae l Eixc4t 28 .. <t!ib1 �b6 29.'!Wa3 Eic8 Black has some compensation for the material. 2S.exdS hdS

25 . . . 'lMff2t 26 . .ie2 .ixd5 27. Eiafl 'lMfc5 28. <t!ib1 �xc4 29.�xc4 �xc4 30 . .ixc4 Eixc4 3 1 . Eid1 and White wins. 26.tlYxdS gd8

Better was 26 . . . �f2t 27. <t!ic3 'lMfe3t 28. 'IMf d3 �b6 Now White should play very carefully. 29. Eid1 e4 30.'lMfc2! .if6t (30 . . . 'lMfe3t 3 1 . <t!ib4 �c5t 32. <t!ia4 'lMfc6t 33. <t!ia3 .ie7t 34.b4 Eib8 35 . � c3t <t!ig8 36.c5 and White wins.) 3 1 . <t!id2 exf3 (3 l . . ..ig5t 32. <t!ie1 exf3 33.'lMfc3t .if6 34.�xf3 �b4t 35 . <t!if2 �xb2t 36 . .ie2 and White should win.) 32. <t!icl and White wins. 27.tlYcS lld2t 28.<bb3 <bg7 29.a3 e4 30J�ih3 tlYf4 31 .fxe4 1-0

228 The Four Knights

The Four Knights

- By Alexander Raetsky

1.e4 c5 2.�f3 �c6 3.d4 cxd4 4.�xd4 �f6 5.�c3 e6 6.�db5 J.b4

The Four Knights variation of the Sicilian, as seen in the diagram above, is a rather sad variation that I , for reasons that will soon be clear to everyone, have chosen as my main defence. Some years ago 1 wrote a book called Meeting 1.e4 for Everyman. For this book my editors, Jacob Aagaard and Byron Jacobs, decided that the Four Knights variation was a timely choice: despite seeming quite playable it had found no real attention in recent theoretical works. Since 1 prefer to play football with my son rather than study chess, this became the line 1 knew best from Black's point of view, and therefore 1 chose to adopt it in my own games.

1 have now agreed to write another article about this line, this time with a few recommendations for White, on the condition that my jokes would not be removed from this manuscript. (I lied - Aagaard)

So let us move to the few lines you need to know to be able to get an advantage against me with White - if I still play this rubbish when this book comes out.

Game 53 Timoshchenko - Chemov Bucharest 1 993

1 .e4 c5 2.�f3 �c6 3.d4 cxd4 4.�xd4 �f6 5.�c3 e6 6.�db5 J.b4 7.a3

7. !f4!? is a sharper line. However, 1

see no reason why it should be stronger, or even provide White with a guaranteed advantage. So 1 recommend just snatching the two bishops and getting on with it. 7 . . . J.xc3t 8.�xc3 d5 9.exd5 exd5

9 . . . lihd5 is a less popular alternative in this position. As in all other lines White should be slightly better. Here 1 will give two examples, but basically there is not a lot of theory here, nor a lot needed, as the position is more technical than tactical in nature.

a) 1 O .tiJxd5 �xd5 1 1 .�xd5 exd5 1 2.!f4 !f5 1 3 .0-0-0 0-0-0 14.f3 h5 1 5J:1d2 E:he8 1 6.!g3 a6 17.!e2 E:e6 1 8 . E:hdl E:d7 19 .!f1 d4 20.!f2 E:c7 2 1 . !d3 !xd3 22. E:xd3 E:e2 23. E: ldl E:xd2 24. @xdl E:d7;!; Peng - L.


Christiansen, Yerevan 1 996. b) 10 .id2!? lLlxc3. The safer choice.

( 10 . . . 0-0 1 1 .�h5!? lLlxc3 ( l l . . . lLl f6 12. �h4 � d4 would still have kept White's advantage to a minimum. 1 3 .�xd4 lLlxd4 14 .0-0-0 e5 1 5 . lLl b5 lLlxb5 1 6.ixb5 if5 17.f3;!;) 12 .ixc3 e5 13 .id3 g6 14 .�h6 �f6 1 5 .0-0 �g7 16 .�e3 g5 17.h4!? h6 1 8 .hxg5 hxg5 1 9 .f3 ie6 20. i>f2± with an attack in Tal - Liberzon, Kislovodsk 1 964.) I l .ixc3 �xdl t 12, lhdl f6 13 .f4 id7 14.ic4 0-0-0 1 5 .0-0 i>c7 1 6J �de 1 l'!he8 In�f3 ic8 1 8 J �g3 l'!e7 19 .1'!ge3 l'!d6 20.b4;!; Fischer - Addison, New York 1 962.

These kind of slightly worse endgames for some reason appeal very little to Black players, while being slightly worse with the queens on is another story. lO . .id3 0-0 1 1 .0-0

This is the standard position in this system. In the next game we shall consider one of the main moves, 1 1 . . . d4, here we

look at two alternatives. 1 1 . . . h6

1 1 . . .ig4 12.8. Now we have two alternatives:

a) 12 . . . ie6 13 .ig5 �b6t 14. i>hl lLld7 Only move. ( 14 . . . �xb2? loses directly to a very simple attack. 1 5 .�d2 �b6 1 6.ixf6 gxf6 17.�h6 f5 18 .lLlxd5! ixd5 19 .ixf5 and Black cannot defend himself) 1 5 .f4 f5 ( l 5 . . . h6? 16.ixh6 gxh6 17.f5 lLlde5 1 8 .�d2 lLlxd3 1 9 .�xh6 would allow White to establish a winning attack. Black can only escape with 1 9 . . . �d4 20.cxd3 �g7 when White will win all the same. Strongest is probably 2 1 .�h4!? ) 1 6.b4 d4 17. lLl a4 �c7 18 . l'!e1 id5 ( l8 . . . l'!fe8;!; was better.) 1 9 .c4! if7 20.ixf5 h6 2 1 .ixh6! ixc4 22.ie6t ixe6 23. l'!xe6 lLlf6 24.ig5 �f7 25. lLlc5 and White won in Tseshkovsky - Maiorov, Novorossijsk 1 995 .

b) 12 . . . ih5 13 .ig5 �b6t 14. i>hl lLl e4 It i s hard to see any other good moves for Black. 1 5 .lLlxe4 dxe4 16.ixe4 �xb2 17 .�bl ( l 7.�d5!? ig6 18 . l'!abl �xa3 1 9 .ixg6 hxg6 20. l'!xb7 �a6 2 1 . l'!fb l;!;) 17 . . . �xb l 18 . l'!fxbl f5 19 .id3 ( l9 .id5t if7 20.c4;!; was also interesting) 1 9 . . . b6 20. l'!b5 ig6 2 1 . l'!d5 h6 (2 1 . . . f4 22.ixg6 hxg6 23. l'!el l'!ae8 24Jhe8 l'!xe8 25 . i>gl B:e2 26.ixf4 l'!xc2 27. l'!d6;!;) 22.ih4 i>h7 23. l'!el l'!ae8 24. l'!xe8 l'!xe8 25.f4 l'!f8 26. i>gl;!;. White retains some pressure in the endgame, Tseshkovsky - Barlov, Budva 1 996. All in all a sad story for Black.


12 . .if4 d4 12 . . . a6 has also been tried. In

Areshchenko - G. Kuzmin, Kramatorsk 2003, White replied naturally with 13 J�el d4 14. 1t1e4 and now Black went wrong with 14 . . . ifS?! ( l4 . . . 1t1xe4 1 5 .ixe4 �f6 1 6.ig3;!; was preferable) when White had a tactical shot with 1 5 .ic7! �xc7 16 .1t1xf6t gxf6 17.ixf5 �f4 1 B .ie4 gadB 1 9.93 �g5 20.h4 �g7 2 1 .�h5 ± . 13.1t1b5!

Knight sorties to other squares promise little according to common practice. 13 ... 1t1d5 14.Wf3!?

14.ig3 ie6 1 5 . gel �d7 1 6.h3 gadB 17 .�f3 ltlde7 1B .ltld6 id5 1 9 .�h5 fSec Krogius - Tal, Riga 1 95B. 14 . . .J.e6

14 . . . 1t1xf4 1 5 .�xf4 leaves the d-pawn in trouble, and after 1 5 . . . ie6?! ( l 5 . . . a6 1 6. 1t1 d6;!; is necessary, though unpleasant) White has 16 . ltl c7! gcB 17 .�e4 ifS 1 B .�xfS g6 1 9 . 1t1 e6!± . 15JUel

15 . . . Wd7?! This is too automatic as will soon

be apparent. Black had two preferable alternatives.

1 5 . . . �f6 16 .�e4 ifS 17.�xfS ( l7.Wxd5 ixd3 I B .id6 ixc2 1 9 .ixfB gxf8 offers Black counterplay with the dpawn) 17 . . . �xfS I B .ixfS ltlxf4 1 9. gadl gadB 20.ie4 and White i s better.

1 5 . . . a6! ? 1 6. 1t1 d6 ltlxf4 17 .�xf4 �c7 I B .�e4 g6 1 9 . 1t1 c4 gfeB 20.�h4 i'g7 and White is only slightly better. 16.ie5 �xe5 17.gxe5 �e7?!

I have had this position once. I played 17 . . . ltl f6 when after l B . �f4 gfdB l9 . gdl ig4 (Bromann - Raetsky, Taastrup 2002) my opponent should have played 20.f3! when the problems with the d-pawn leave me clearly worse. e.g. 20 . . . ie6 2 1 . if1 ± . Instead my opponent played something else and we drew 1 36 moves later. 18.We4 :afd8?!

Black decides not to let the pawn go. It was a sad choice. He could have kept his kingside position together with 1 B . . . ltl g6 when White wins a pawn with 1 9. 9c5 ( l 9. gxe6!?) 19 . . . gfcB 20. gxcBt gxcB 2 1 . 1t1xd4± and he should win. 19.Wh7t �f8 20.Wh8t �g8 21 .:aael .id5 22J::lxd5!

The bishop must be stopped from coming to e6. Now Black is mated. 22 .. . Wxd5 23.�c7 Wd7 24.ih7 1-0


Game 54 Karjakin - Raetsky Biel 2003

1 .e4 c5 2.�6 �c6 3.d4 exd4 4.�xd4 �f6 5.�c3 e6 6.�db5 1M 7.a3 ixc3t 8.�xc3 d5 9.exd5 exd5 lo.1d3 0-0 1 1 .0-0 d4 12.�e2

1 2. tD e4 might seem to be the natural option, but it leads to simplification and boring positions after 1 2 . . . .tf5 1 3 . .tg5 .txe4 14 . .txe4 h6. 12 ... 1g4

1 2 . . .'� d5 is the main alternative to 12 . . . .tg4. 1 3.c4 White should try to open the game and use the strength of the two bishops. (White had a slight advantage after 1 3 .tDf4 �d6 14 . tDh5 tDxh5 1 5 .�xh5 h6 16 . .td2 .td7 1 7J'1ae l !'1feS I S .f4 tD e7 1 9. !'1e5 .tc6 20 . .tb4 �f6 2 1 . .tc4 Xie - Z. Polgar, Jaen (S) 1 996) 13 . . . dxc3 14 . tDxc3 �a5 1 5 . .tf4 .tg4 16 .�bl tD d4 1 7.b4 �h5 I S .f3 .te6 19 .�b2 !'1fdS 20. !'1aeU/=, Kamsky -Lobron, Dortmund 1 993.

12 . . . !'1eS is another quite solid alternative. 1 3 . .tg5 !'1e5 ( 1 3 . . . h6 14 . .th4 a6?! (Better was 14 . . . g5 1 5 . .tg3 tD e4 16.f3 tDxg3 17 . tDxg3 tD e5 and Black is only very slightly worse) 1 5 . !'1e 1 g5 16 . .tg3 tD e4?! ( 1 6 . . . tDh5 was better) 17. tD xd4! �xd4 I S .c3 .tg4 1 9 .cxd4 .txdl 20 . .txe4 .tb3 2 1 .d5 tD d4 22.d6± with a close to winning ending for White in Alekseev - Ianocichin, Oropesa del Mar 200 1 .) 14.f4!? !'1d5 1 5 . tD g3 ( 1 5 .f5! ?

with various attacking ideas also looks good) 1 5 . . . h6 1 6 . .th4 g5 ! ? Otherwise White is just a whole lot better. 17.fxg5 hxg5 I S . tD h5! Only move. I S . . . tDg4 1 9 . .tg3 tDe3 20. tD f6t �xf6 2 1 . !'1xf6 tDxdl 22. !'1xdl with a better endgame for White in Korneev-Moreno, Mondariz 2000. 13.6

Harmless is 1 3 . .tg5 �d6 14 .!'1el !'1feS 1 5 .�d2 .txe2! The standard exchange in this variation. 1 6 . .tf4 �d7 17 . !'1xe2 !'1xe2 I S .�xe2 !'1eS 19 .�f1 �e6 20.h3 h6 2 1 . !'1dl tD d5 22 . .tg3 tDf6= Kasparov - Grischuk, Cannes 200 1 . 13 .. .ih5

14.M!? White takes risks by allowing the

weaknesses at first e3, and now the c3 and c4 squares. Obviously Black has some counterplay here, but White avoids simplifications with the strategy of winning the isolated d4-pawn, and using


the strength of his bishop pair. This is a typical modern chess struggle.

In my games I have also faced 14. ttl f4 �g6 l S . ttl xg6 hxg6 1 6.f4 ttl e7 (also possible is 1 6 . . . ttldS 17.fS gxf5 l S. E1xf5 ttl e3 1 9 .�xe3 dxe3 when 20.�hS!? g6 2 1 . �h6 �d4 22.E1hS!? gives White a very interesting attack.) 1 7. E1e 1 ttl fS l S .�£3 �d7 19 . E1eS a6 20.!d2 E1feS 2 1 . E1ae 1 and White is definitely better, A. Sokolov - Raetsky, Basel 2003. 14 ... YNb6 15.�f4 .tg6 16.�xg6 hxg6 17.f4

17 ... �d5 White should be better no matter

what. Another example is: 17 . . . � e7 I S . �f3 ttl edS 19 .E1e l E1feS 20.!d2 �c7 (20 . . . ttl e3 2 1 .c4! is generally good for White) 2 U � {h1 �d6 22.fS gxfS 23.�xf5 g6 24.�d3 I!fg7 2S .h3 E1hS 26. E1f1 E1aeS 27.E1ae 1 with an advantage for White in Galkin - Rabiega, Ohrid 200 1 . 18.We

This is probably best. l S .fS ttl e3 1 9 .!xe3 dxe3 20. l!fh1 (A

bad direction is 20.�e2 gxf5 2 1 . !xf5?! 2 1 . . . ttl d4 22. � xe3? when Black wins with 22 . . . E1aeS 23.�£2 E1e2) 20 . . . ttl eS ! ? (20 . . . gxf5 2 1 . E1xf5oo) 2 1 .fxg6 fxg6 and Black has good counterplay. 18 ... �e3

19JUl! Subtle play from the kid who recently

described his greatest fear in life as not becoming World Champion. If he takes a close look at what happened to his friend Ponomariov, he should maybe fear becoming World Champion more than not doing so!?

Anyway, 1 9 .E1e l E1feS 20.!d2 looks natural, and the computer thinks White is better, but Black has 20 . . . ttlxc2! 2 1 .!xc2 d3t 22.�£2 (22.!e3 tDd4 23.�£2 E1xe3! and Black ends on top.) 22 . . . E1xe l t 23. E1xe1 dxc2 24.!e3 �a6 and White is struggling to keep equality.


19 •. .l:i:£e8 20 . .id2 ge7 Here I could also have continued 20 . . .

a5 !? White i s still better after 2 1 .b5 f:iJe7 22Jl:e1 f:iJ 7d5 23 .:1l:fe2 :1l:acS 24.c4 dxc3 25 . .!xe3 :1l:xe3 26. :1l:xe3 c2 27 . .!xc2 f:iJxe3 2S:�xe3 �xe3t 29. :1l:xe3 :1l:xc2 30. :1l:eSt �h7 3 1 . :1l:e7 and White has some winning chances in the endgame, even though I have drawing chances too! 21 .h4!? :Bae8 22.h5 p!?

He wants to complicate the game and I am not afraid to follow suit. I estimated that after 22 . . . gxh5 23.�xh5 g6 24.�g5 White would have the better game. Black should fear the advance f4-f5, and the white rook will also come into play by f3-h3 with dangerous threats.

23.h6! OK, the little guy is not so bad. My

main idea was 23.fxg5 f:iJe5 24.�e4 (24.'i;lfg3 f:iJxd3 25.cxd3 'i;lfb5 and I cannot for the life of me pretend to be worse here.) 24 . . . f:iJxd3 25 .'i;lfxd3 :1l:e5

26.g6 fxg6 27.hxg6 :1l:Se6 and I think I am fully in the game. 23 ... g4

I considered 23 . . . gxh6 24. 'i;lfh5 to be very uncomfortable. 24.Wg3

24.h7t �hS 25 .'i;lfg3 is not clear at all. Maybe White is better, but I am not sure about that at all. 24 ••. g6?!

Unfortunately I was running short of time. My alternatives here were 24 . . . f5? 25.hxg7 :1l:xg7 (25 . . . �xg7 26 . .!xf5! is not right.) 26. :1l:e2 :1l:ge7 27. :1l:ae1 and I would simply lose a pawn without compensation. 24 . . . gxh6! was the right choice now. After 25 .'i;lfh4 �g7 26. :1l:e1 White still has some pressure, but I would have had good chances of neutralising it. 25.b5! �b8 26.£5

26. 'i;lfh4!? also looked quite strong. After 26 . . . :1l:e6 27. :1l:el White is threatening f5 and after 27 . . . f5 2S. :1l:fe2 he is contemplating c3 . White is better, but how much? 26 ... �d7?

This is just a mistake. Necessary was the uncomfortable 26 . . . gxf5 opening up the king. White is better, but at least there is some limit to how much. 27 • .ib4

Ouch. 27 .• J3e5

27 . . . f:iJ c5 2S .'i;lfh4! and there is no real defence against .!xc5 and 'i;lf f6 as 2S . . . :1l:e5, with the idea of later playing


. . . �f8, is met by 29.fxg6 fxg6 30J'�f6 with a winning attack. 28.fxg6 fxg6 29.YlYf4

Now it is all over. However I play on . . . 29 ••• YlYe6 30.YlYxd4 �h7 31 .Ad6 g3 32.ge2 �5 33.ixf5 YlYx5 34.be5 �xe5 35J::fael YlYh5 36J::fxe5 YlYh2t 37.�f1 gmt 38.�e2 YlYxg2t 39.�dl 1-0

The Cobra Variation

1 .e4 c5 2.�a e6 3.d4 cxd4 4.�xd4 �f6 5.�c3 �c6 6.�db5 J.c5!?

This line is called the Cobra and was popularised by an article in New In Chess by Swedish 1M Jesper Hall who, among other games, presented his win against grandmaster Krasenkow. However, since then White has been able to locate a simple way to gain a slight edge. This has not prevented me from playing it way too ofren!

7.J.f4! This is the simple choice. If White plays

like this only masochists will continue to play the Cobra with Black. 7 .•• 0-0

7 . . . e5 is met strongly with S . .te3! d6 (S . . . .txe3? 9. lLld6t r,!;>f8 1O.fxe3 �b6 1 1 .�d2 �xb2 12 .gb1 �a3 13 . .tc4±) 9 . .txc5 dxc5 10 . �xdSt r,!;>xdS 1 1 .0-0-0t with a very nice endgame for White. 8.J.c7

Also interesting is S.e5!? lLl eS 9 . lLl e4 .te7 1 0.c3 ! ? when White is probably better as well. 1 O . . . f6 1 1 .exf6 .txf6 ( 1 1 . . .lLlxf6?! 1 2 . .tc7 �eS 13 .lLlxf6t gxf6 14 . .tg3 �f8 1 5 . lLl d6 100ks very good for White.) 1 2 . .td6 lLlxd6 1 3 .lLlbxd6 .te5 14 . .te2 Y!1c7 1 5 .�d3 lLle7 16 . gd1 lLl d5 1 7. lLl g5 lLl f6 1 s . lLl f3 .!xd6 1 9.Y!1xd6 Y!1xd6 20. gxd6 r,!;>f7 and the endgame slightly favours White - Hall. 8 ••• YlYe7 9.J.d6 J.xd6 lO.YlYxd6

Basically this position is depressing


and passive for Black. All his dreams of winning the game seem to have gone. 10 • • .tl�e8

1 O . . . �d8 1 1 .0-0-0 a6 12. ttJ d4 �b6 13 .ttJxc6 bxc6 { 1 3 ... dxc6 14 .Eid2 �a5 1 5 .f4 b5 1 6.�c5 also favours White. Soon will come e5 and ttJe4-d6.) 14.£3 a5 1 5 . ttJ a4 �b4 1 6.b3 �xd6 17 . Eixd6 ttJe8 1 8 . Eid3 d5 19 . Eic3 ttJf6 20.id3 Eid8 2 1 . Eie 1 i.b7 22. ttJb6 Eia7 23. Eic5 ;!; Oll - Janssen, Dieren 1 996.

Even more passive is 10 . . . �xd6 l 1 . ttJxd6 ttJ e8 12 .ttJxe8 Eixe8 13 .0-0-0 Eie7 14. ttJ b5 b6 1 5 . ttJ d6 Eib8 1 6.i.b5 mfB 17 .i.xc6 dxc6 1 8 .ttJxc8 Eixc8 1 9. Eid6, which left Black a miserable endgame in Aagaard - Bellon Lopez, Gothenburg 2004. 1 1 .'i!Yxe7 tLlxe7 12.0-0-0 f5 13.tLld6

The simplest. Also possible is 1 3 .e5 ttJg6 14 .Eie1 a6 1 5 . ttJ d4 b5 16 .g3 i.b7 17. Eig1 f4 1 8 .id3 with a slight edge for White, or 1 3 .£3!? fxe4 14.fxe4 a6 1 5 .ttJd6 ttJxd6 1 6. Eixd6 b5 1 7.i.e2 ttJg6 18 . Eib6 ttJ e5 19 . Eifl Eixfl t 20.ixfl and White had some advantage in Fichtl -Opocensky, Prague 1 953 . 13 ••• tLlxd6 14J:gxd6 fxe4 15.tLlxe4 tLlf5 16J:!d2 d5 17.tLlg5 b6!?

This is a marginal improvement. 1 7 . . . id7 1 8 .i.d3 e5 is met strongly with 19 .i.e4!? dxe4 20. Eixd7 e3 2 1 .fxe3 ttJxe3 22. Eie1 ttJxg2 23. Eixe5 Eiad8 24. Eixd8 Eixd8 25 . Eie7 ttJf4 26.b4 and White has good winning chances - Hall. 18.tLla tLld6 19.i.d3 .id7 2oJleI;!; Berg - Hall, Germany 2002.

This kind of endgame often arises from the French. White can now choose between different plans. The main problem for Black will always be his slightly worse bishop and pawn structure. He was able to draw the game with a pawn sacrifice and tenacious defence, but it was a difficult and unpleasant task.

236 The Pin Variation

The Pin Variation

- By Jacob Aagaard

1 .e4 c5 2.ttla e6 3.d4 cxd4 4.ttlxd4 ttlf6 5.ttlc3 1b4

This rare and provocative line is generally considered unsatisfactory for Black, as White will be able to start a kingside attack with the support of his advanced e-pawn. However things are not that simple. In recent years players such as Federov, van Wely and in particular the Lithuanian grandmaster Sulskis (a legendary fighter on the European tournament circuit, as well as a really nice guy) have employed this system with relative success. For the 1 .e4 player who brakes for no one, insight into this line seems to be becoming more and more important. For this reason we have decided to expand from the planned 2 pages to a full size chapter.

However, an interesting alternative is to play 3 . l2k3!? in reply to 2 . . . e6. Now after 3 . . . Nf6 we can transpose to the Nimzowitsch variation with 4.e5! , as dealt with on page 255. And after all other reasonable moves 4.d4 leads into the normal main lines. In this way there is also a link to the Kan/Taimanov lines, where after 2 . ttJ f3 e6 Sune Berg Hansen suggests 3. ttJ c3!? as a viable alternative. This actually prevents the Pin-variation; but who wants to do that?

The following line is the established refiltation of the most important sideline with 6.e5 ttJ e4? ! , which is dose to a losing mistake, though it has taken decades to prove it! In the sidelines White wins, but only after several minor improvements.

l .e4 c5 2.ttla e6 3.d4 cxd4 4.ttlxd4 ttlf6 5.ttlc3 1b4 6.e5 ttle4?!

6 . . . . �a5 ? can be refuted in more than one way, but the following is probably simplest. 7.exf6 .ixc3t B .bxc3 �xc3t 9 .�d2 �xal 10 .c3! �bl 1 1 ..id3 �b6 1 2.fxg7 The only debatable move, but here we just go for it. 12 .. J'1gB 13 .�g5! Here White gets an endgame with two bishops and 2-3 passed pawns, where at least one is extra. 1 3 . . . h6 ( 1 3 . . . �dB 14 .�xdBt mxdB 15 . .ig5t mc7 16 . .if6 d6 17 . .ixh7 ttJd7 I B . .ixgB ttJxf6 1 9 . .ixf7+- Ngo Ngoc - Bao, Dalat City 2004.) 14 .�f6 1];\IdB 1 5 .�xdBt! Always this. 1 5 . . . mxdB 1 6 . .ixh6 f5 17 . ttJxf5 exf5 I B . .ic4 Ei:eBt 1 9 . mdl ttJ c6 20 . .if7


d6 2 U ! :e 1 lLl eS 22.f4 and White won in Kozakov - Todorov, Valjevo 2000. 7.YlYg4 ti)xc3

Again 7 . . . YlY as should be a losing move. B .'IWxg7! (B.�xe4!? !xc3t 9.bxc3 �xc3t 10 . �d1 'lWxal l 1 . lLl bS dS is less clear than I want it to be. Probably it also wins though, but the main line seems most convincing to me.) B o o .!xc3t 9.bxc3 'lWxc3t 10 . �e2 b6 l 1 .'lWxhBt �e7 12 .!a3t! when White wins. Most sources stop here, but let us look a few moves further. 12 o o .'lWxa3!? ( 1 2o o .d6 1 3 . lLl b3! wins easily for White. There is no reason for 1 3 .lLlc6t?! as played in some games.) 13 . 'lWxcB 'lWb2

14. lLl b3 lLlc6 1 S .'lWxaB 'lWxc2t 1 6. �e3 'lWxf2t 17. �xe4 lLlxeS 1 B .�f8t!! ( 1 B .�xeS �e3t 19 .'lWe4 f6 mate is wonderful, and probably blinded Black.) 1 B o o . �xf8 1 9 . �xeS 'lWe3t 20. �d6 �eB 2 1 . �c7 1-0. Schatzle - Filartiga, corr. 1 974.

This is the line you could expect to find elsewhere, as the game is certainly very nice. But instead of 14 . . . lLlc6?? Black has 14 o o .�xeS, which is a winning novelty. After l S .f4 �xf4 there is no defence. One line is 1 6.c4 'lWf2t 17 . �d3 lLlc6 1 B .'lWxaB lLlb4t 19 . �xe4 fSt 20. �eS d6 mate.

So White needs to come up with 14 . �e3! , which is actually the only winning move here, and still it is not easy.

a) 14 o o . lLlxf2 l S .!e2 lLlxh l 16 .Eidl ! (The most accurate move. White is threatening 1 7. lLl c6t!) 1 6o o .�a3t 17. lLl b3 'lWb2 1 B .c3 and White wins.

b) 14 o o .�xa1 l S . �xe4 �e1 t 16 .�f3 'lWxeS 17 . �b7 YlYxd4 l B . 'lWxaB lLl c6 1 9 .h4! and the white king will escape. However this win is not hard to find. On move 16 there might be alternative ways to do it, but here 1 9 .�gB? lLleSt 20. �g3 lLl g6! ! would draw for Black. 8.YlYxg7

When I used to play the Pin Variation in Blitz I would often lose to B.a3, but after B o o .!f8! Black seems to be OK. 8 •• J�fB 9.a3 ti)b5t

Black has the following alternatives: 9 o o . lLl c6 10 .axb4 lLlxd4 l 1 .bxc3 lLlxc2t

12 . �d1 lLlxa1 , which simply loses to 1 3 .!gS ! as in Mendoza - Blejman, Guaymallen 200 1 among others. Now 1 3 o o .�xgS 14 .YlYxgS as is the computer's idea. The word desperation springs to mind.


9 . . . .ia5 1 O . .ih6 �e7 I l . itl b3! and I cannot see any justification for the exchange Black will eventually lose.

9 . . . � a5 is the most complicated move here, but Black will not escape after some simple moves from White. 1 0. itl b3 V;l/d5 ( l O . . . itl e4t I l .axb4 �xb4t 1 2.c3 itlxc3 13 . .ih6 itle4t 14 . itl d2 itlxd2 1 5 . V;l/xfBt V;l/xfB 16 . .ixfB itl b3 17J '!:a3 and White won in Elis - Schork, Bonn 1 998.) 1 1 ..id3 itl a2t ( l 1 . . ..ie7 12 .bxc3 Schmidt - Boidman, Bad Breisig 2000.) 1 2.axb4 itlxc1 1 3 J�xc1 ± Harasimovic -Berisha, Brno 1 997. 10.axb4 itlxd4 l l..tgS Wb6

The standard move. 1 1 . . .itlxc2t 12 . @dl �b6 was successful for Black in May - Beier, East Germany 1 979, but here 1 3 . .ih6! V;l/xb4 14 . @xc2 would lead to an endgame where White has an extra exchange against a pawn, and every chance of winning. 12 . .ih6

12 . .id3 was successful in some games, but 1 2 . . . d6 13 .c3 dxe5! leaves the position unclear. 12 ... Wxb4t 13.d �f5 14.cxb4 �xg7 15.bg7 ggS 16 . .if6:t

This is the old main line. Now 16 ... �c6!?

was analysed by my friend Torben Sorensen in Denmark a long time ago. The conclusion was the following wonderful winning line. Alternatively Black can try 1 6 . . . d5 to fight for a draw in a very depressing way. After 17.exd6

itl d7 1 8 . .ic3 itlb6 as in Krumova -Teodorescu, Bydgoszcz 1 978, I think the simplest is 19 .93 @d7 20 . .id3 f5 2 1 ..ie5 ± . Is White not just a pawn up?

17.ga3! This wonderful manoeuvre was Torben's

recipe against his own idea. 17 ... �xb4 1S.gh3 �d5 19J�xh7 �xf6

19 . . . d6! ? is better, but White has the advantage after 20.h4! with similar ideas. 20.exf6 d5 21 .h4

2 1 . .ib5t? only helps Black to play the move he wants to play, . . . @ffi with the idea of preventing E:g8t. However if White plays accurately, Black is too late with his planned counterplay on the queenside. 21 ... �f8 22.h5 id7 23.h6 gcs 24J�g7 ghS 25.h7 ia4 26JlgSt! f:1xgS 27.hSW! and White wins.

So 6 . . . itl e4 is discredited, but the same does not go for 6 . . . itl d5 .


The Pin-variation with 8 •• .Aa5.

1 .e4 c5 2.tiH3 e6 3.d4 cx:d4 4.�xd4 �f6 5.�c3 1b4 6.e5 �d5 7.id2

Experience has proved that this is the strongest move here. If you have too much time on your hands you can try to investigate 7. 'W g4 and see if this is really true. Here I will focus on the main line. 7 ... �xc3 8.bxc3 i.a5!?

The bishop abandons the defence of the kingside. However it is not so easy to refute, and a clear refutation has never occurred over the board. The bishop does leave the kingside unprotected, but on the other hand the bishop is not a very good defender in many lines, and now leaves room for the queen. 9.�g4 0-0 lO.i.d3!

This natural developing move has to be the right move. 10 .�g3!? has scored well for White, but no one has played the principled 1 0 . . . f6! when the position is

far from clear. Why should Black not be fine here?

Another way to play the opening for White is 9. 1d3 when 9 . . . 0-0 10 .�h5! looks nothing like the transposition which could occur after 10 . 'W g4. However Black has other possibilities. 9 . . . 'W c7 1 0.0-0 a6 . Being the only move that does not seem to lose by force, this is of course a new move. 1 1 . 'W g4 g6 1 2. � g5 1xc3 13 .1xc3 'Wxc3 14 .'lWf6 0-0 1 5 . � e2 'Wc5 1 6. ttJ g3 d6 17.exd6 'Wxd6 I B .ttJh5 gxh5 1 9.'Wg5t Y2-Y2 was played in R. Jones - Aagaard, England 2004. It has not been possible for me to prove an advantage for White in this game, despite not believing much in my own play. lO ••• d6!

This is, of course, the way Black should play. Now it gets really interesting. The alternative is rather passive: 1 0 . . . ttJ c6 l 1 . ttJxc6 dxc6 12.0-0 'lWd5 13 .'lWh4 g6 14 .1h6 1dB 1 5 .'lWf4 EieB 16 . EifeH . . This position i s very uncomfortable for Black. Now the game finished 1 6 . . . 'lWa5 17. Eie3 'Wxc3 I B . Eidl 'lWc5 19 .Eif3 'lWe7 20.1e4 1b6 2 1 . 1g5 'Wf8 22.'Wh4 h5 23.g4 'lWg7 24.gxh5 'Wh7 25.1f6 1-0. Wosch - Rueppel, e-mail 200 1 . 1 1 .�a!

The most aggressive. l 1 .exd6 is my computer's preference.

But now 1 1 . . .5! 12 . 'W g3 e5 13 . ttJ b3 1b6 14 .0-0 'lWxd6 1 5 . Eiadl ttJc6 is closer to worse than to better for White. Lanc - Amason, Prague 1 9BO.


The direct attack with 1 1 .!g5!? was analysed by Byrne and Mednis.

The lines credited here to the two American gentlemen are quoted from Nuno's and Gallagher's Beating the Sicilian 3. Essentially Black has two choices:

a) 1 1 . . .1lN c7 1 2.!f6 '?Nxc3t 13 . @e2 And here it is not hard to find out that 1 3 . . . g6! is the superior move order. ( 1 3 . . . VNdlt 14 .@f1 g6 1 5 J�dl ! was the line given by Byrne and Mednis.) 14 .'�g5 tLld7 1 5 .'?Nh6 ( 1 5 . tLl b5 VNc5 1 6.exd6 eH) 15 ... VNd2t 16 .VNxdl !xd2 17 . @xdl dxe5 looks good for Black.

But what about 12 .!xh7t!! @xh7 13 .!f6 VNxc3t 14. @f1 gxf6 1 5 .exf6 VNxal t 1 6. @e2 and White wins? This is beyond the immediate scope of the computers, but not this writer! (nice to have been of assistance Mr. Fritz . . . )

b) 1 1 . . .!xc3t! This move order is probably the most reliable. ( 1 1 . . .f5! ? is also possible, with the idea of transposing,

but why bother?) . 12 . @f1 f5! This is absolutely forced. ( 1 2 . . . VN a5 loses easily to 1 3 .!f6 g6 14 .VNg5 !dl 1 5 .f4 and Black cannot protect his king. And 1 2 . . . f6? fails to 13 .'?Nh5! g6 ( 1 3 . . . h6 14 .VNg6 and White should win.) 14.!xg6 VNd7 ( 14 . . . VNc7 1 5 .!xf6 !xd4 1 6.exd6 VNc4t 1 7.!d3 and White wins the queen for insufficient compensation.) 1 5 . tLlxe6 !xal 1 6.'?Nh6!+-. The attack is too strong) 1 3 .exf6

1 3 . . . '?N d7! . The only move. It looks rather shaky, but to refute it is not so easy. ( 1 3 .. J �xf6 was part of the original analysis by B&M, but now 14 .!xh7t! @xh7 1 5 .!xf6 '?Nxf6 16 .'?Nh3t @g8 17. VNxc3 tLlc6 18 J :1dl and White wins.) 14 J :1bl ( 14 J:1dl g6 1 5 .!b5 VNfl and the position appears unclear to me.) 14 . . . tLl c6! ( 14 . . . g6?! 1 5 J :1b3 !xd4 16 .VNxd4 tLl c6 17. '?Nh4 e5 gives White a strong attack with 1 8 .'?Nh6 VNfl 19 .h4 and the black king is very tender.) 1 5 .fxg7! ( 1 5 .tLlxc6


bxc6 1 6.fxg7 Wixg7 17.lMfh4. White might be better, but how clear is it after 17 .. Jl:f7!. It is possible to analyse these positions until one goes blue in the face. I will stop here saying that the position is complicated, and a clear conclusion is hard to make. Probably the chances are about level.) 1 5 . . . lMfxg7! ( I 5 . . . �xf2t 16. cJixf2 .ixd4t 17. cJie1 Wixg7 1 8 .lMfh5 and White wins. Black never got out) 1 6. lLl xe6 and now:

b I ) 1 6 . . . .ixe6? 17. Wixe6t cJih8 1 8 . .ih6 1Mff6 ( I 8 . . . lMfd4 1 9 . .ixfS �xfS 20.f3 cannot be working) 19 .1Mfxf6t �xf6 20.!g5 ± .

b2) 1 6 . . . lLl e5! 17 .lMfc4!? ( I 7.!xh7t lMfxh7 1 8 .lLlxfS lLlxg4 1 9 .1Llxh7 cJixh7 is just not very clear.) 17 . . . Wif7! ( I7 . . . lLlxc4 1 8 .lLlxg7 lLlb2 1 9 .1Llh5 lLlxd3 20.cxd3;t with chances for White in the endgame, though Black has reasonable drawing chances) 1 8 .!xh7t Wixh7 19 .1LlxfSt lLlxc4 20. lLlxh7 cJixh7 and the endgame could have any of three results.

I think we can conclude that 1 1 ..ig5!? is not a killer, though clearly interesting. l I ...g6?

This is the traditional move, but my analysis seems to suggest that Black cannot allow White to attack the king unhindered like this. Suddenly the critical move becomes 1 1 . . .f5 ! 12 .exf6 �xf6 13 .0-0 h6 14 .1Mf e4;t . I believe that White is better here, but it is not easy to prove. As there are limits to how much independent analysis there is time to provide for a book

like this, especially on the Pin-variation, then I will stop by giving the moves in the game. 14 . . . lLld7!? ( I4 . . . �f5 is also possible, though it appears risky to me not to develop.) 1 5 .lMfh7t cJif7 16.!g5 lLl fS 17 . .ixf6 1Mfxf6 1 8 .Wie4 .ixc3 19 . �ab l d5� Bilbao - Ramos, Alicante 1 989. Is White better in the end here? I am not certain. The game was eventually drawn, but then Black was much higher rated.

On l 1 . . .dxe5?? then 12 . .ixh7t cJixh7 13 .Wih5t cJig8 14 . lLlg5 wins the game. 12.h4!

Again the most aggressive approach is necessary, as White otherwise risks being stuck with a bad structure.

1 2.0-0?! f5! 13 .Wig3 ( I 3.exf6 Wixf6 14 . .ig5 Wig7°o Milosevic - Ammann, Switzerland 1 993) 1 3 . . . dxe5 14 .lLlxe5 1Mff6! ( I4 . . . lLl d7?? 1 5 . lLlxg6+-) 1 5 . .ig5 lMfg7 1 6.!b5 a6 17 . .ia4 !c7 1 8 . �ae l °o This kind of position is close to impossible to analyse. With the aid of a computer we


can give some predictions: 1 8 . . . i.d7 is probably fine for Black is one of them. 12 • • • dxe5

1 2 . . . f5 13 .WI'g3 dxe5 14.h5 transposes. 13.h5 6

1 3 . . . WI'c7 14.hxg6 fxg6 1 5 .0-0-0 leaves White with a very strong attack. I have not found a good defence.

1 5 . . . e4 1 6.�xe4 �xc3 17.i.xc3 Wl'xc3 1 8 . E:xh7! and White's attack is crushing.

1 5 . . . l2k6 16 . E:h6! followed by E:dh l and Black's position will collapse.

1 5 . . . E:f5!? A desperate computer move. 16 . E:h6! There is no reason to take the exchange, even though this also favours White. 1 6 . . . lLld7 17 .WI'h4 lLl fB 1 8 .�xf5 exf5 1 9 .1Llg5 i.e6 2o. lLlxh7 and White wins.

14.Y!Yg3! It is this new move that does the most

damage to Black's position. 14 .i.xf5!? is a very interesting sacrifice,

but not completely conclusive as had

previously been thought. 1 4 . . . exf5 1 5 .WI'c4t Wg7! ( 1 5 . . . E:f7? 1 6.hxg6 hxg6 17 . lLlg5 fic7 1 8 .WI'h4+- Wagman -Barle, Biel 1 98 1 ) 1 6.hxg6 f4! The move overlooked in other sources, which claim the attack is winning. 17 . E:xh7t Wxg6 1 8 .WI'e4t E:f5 19 . E:h3. White has a very strong attack here. Black is probably forced to play 19 . . . Wf7 when White should have more than enough compensation for the piece. But 14 . WI' g3 just wins, so why not play it?

14 .WI'h3?! g5 1 5 .lLlxe5 as in Lerner -Khodos, Rostov 1 976 is less strong. After 1 5 . . . WI'f6!+ I would prefer to be Black. Where is White's attack? 14 ••• e4

14 . . . f4 1 5 .WI'g4 e4 1 6.hxg6! ( 1 6.�xe4 e5 is less clear.) 1 6 . . . exf3 17 . E:xh7 Wl'f6 1 8 .g7 and White wins. 15.hxg6 exf3

1 5 . . . exd3 1 6.gxh7t Wh8 1 7. lLl e5 and White wins.


16J�xh7! This is the way to play the attack. The

h7 -squate is the weakest spot in Black's position, so it needs to be accessible. Again this is a move computers do not find easily, but they really do love it, once they ate told about it.

1 6.gxh7t? would allow the black king to hide behind the white pawn. After 16 . . . @h8 17.'�e5t 'lMff6 ( l7 .. .:!H6!? is maybe safer. After 1 8 . .!g5 lLld7 1 9 . .!xf6t lLlxf6 20.:!�h6! fxg2 2 1 . @e2. White seems to end with a better game, but Black has the clever answer 2 1 . . . .!c7! 22.1Mfxf6t 1Mfxf6 23. l'!xf6 �h2 when the endgame is at best even for White.) 1 8 .1Mfxa5 fxg2 19 .1'!gl b6 20.1Mfb4 '!b7 2 1 .0-0-0 lLldTt I cannot see any reason why this should not be better for Black. An important line is 22.1Mfd4 e5! ! 23.1Mfxd7 .!c6 24.1Mfxf5 1Mfxf5 25 . .!xf5 l'!xf5+ . 16 • • :i!Yf6

There ate no alternatives. 1 6 . . . lLld7 17.0-0-0 lLl f6 1 8 .'lMfh3 and White wins. 17.0-0-0 bc3 18.g7! and White wins.

In the next game we shall see a Danish GM, and co-author of this book, display his full mastery over the opening, by disposing of the old main line of 8 . . . .!e7 9 .1Mf g4 O-O? ! , before we turn to the critical lines of 9 . . . Kf8! and 8 . . . .!ffi!?

Game 55 Hansen - Kristensen Taastrup 1 998

1 .e4 c5 2.�f3 e6 3.d4 cxd4 4.�xd4 �f6 5.�c3 J.b4 6.e5 �d5 7.J.d2 �xc3 8.bxc3 J.e7 9.YlYg4

This is the way to go. 9 • • • 0-0?!

This used to be the main move, but now it has been close to refuted.

9 . . . g6?! is probably too passive and weakening at the same time. I like the true computer move 1 O. lLl b5!? ( l 0.h4 h5 1 1 . 'IMf g3 also looks reasonable.) 1 o . . . lLl c6 1 1 . lLl d6t .!xd6 1 2.exd6 1Mfa5 13 .'lMff4 f5 14 .'!d3± Shredder - Hiarcs 7.32, Debrecen 2000.

9 . . . g5 also does not seem to work. 1 0.h4! h5 1 1 .hxg5 hxg4 12 . l'!xh8t �ffi

wasplayed in Grosar-DeWaal, Belgium 1 986, and now 1 3 .g6 fxg6 14 . .!h6 @f7 1 5 . l'!xf8t 1Mfxf8 1 6 . .!xf8 @xf8 17 . .!e2 is


good for White according to Nunn and Gallagher. However this line is wishful thinking. 14 . . . 'it>f7 makes little sense, and especially 1 3 . . . fxg6 is an automatic move that serves the interests of the opponent. After 1 3 . . . 'W c7! Black is simply winning. The lines could continue 14.0-0-0 'WxeS I S J' !g8 'it>e7 and, though the position is still complicated, it is very hard to believe that White will prove real counterplay.

Instead of l 1 .hxgS? then 1 1 . 'W g3! was the right move. Black has little pleasure in the position. The weakness of the dark squares is terminal. After l l . . .gxh4 12 .'Wg7 l:'i:fB 1 3 .'tJbS White wins. Eventually 'tJd6t and �h6 will ruin Black's position. Necessary is 1 1 . . . g4 12. 'tJ bS a6 ( 12 . . . 'tJc6 13 .�gS ±) 1 3 .'tJd6t !xd6 14.exd6 'tJc6 I S .c4 where I would not mind being White. Black will have a nightmare finding safery for his king, after White develops and then starts to open up the position. 10 . .ih6 g6 1 1 .h4! 'Wa5

Though this is bad, 1 1 . . .d6?! is an even worse move order for Black. After 12.hS 'WaS ( 12. . . dxeS 13 .hxg6 fxg6 14 .!xf8! wins for White, as if Black recaptures on fB then I S . 'tJxe6 and 16 .!c4 is winning. Now 14 . . . exd4 I S .!xe7 'Wxe7 16 .!d3 !d7 17 .l:'i:bl b6 1 8 .cxd4+- was played in Klovans - Schein, Graz 1 999.) 1 3 .'tJbS! Now White has won a tempo compared to the main line, which is also winning. 1 3 . . . a6 14.hxg6 fxg6 I S .!xfB axbS 16 .'it>d2 !xf8 17 . l:'i:xh7! 1-0. B .

Lalic - Sulava, Pula 1 997. White's attack is conclusive: 17 . . . 'it>xh7 18 .�d3 �h6t 1 9.f4 'it>g8 20.'Wxg6t !g7 2 1 . 'We8t !fB 22.!h7t and mate follows. 12.Wg3 d6

12 . . . l:'i:d8 looks entirely wrong, when White is not planning to take on fB at the first given moment anyway. However, it does make sense to play it sooner rather than miss out on it later. But after 1 3 .hS d6 14.hxg6 fxg6 I S .�f4 (keeping the centre stable and the black king exposed) I S . . . dxeS 1 6.�xeS l:'i:dS 1 7.f4 'tJd7 1 8 .!c4! 'tJxeS ( 1 8 . . . l:'i:xeSt 1 9 .fxeS 'WxeSt 20.'WxeS 'tJxeS 2 1 .�b3± - Nunn & Gallagher) 1 9 .!xdS 'WxdS 20.fxeS 'W e4 t 2 1 . 'it> d2 White is completely winning, Wedberg Pokojowczyk, Copenhagen 1 984.

13.�b5! White's strategy in this game is little short

of amazing. Our very own expert in the Kan

and Taimanov seems to bust this system.


13 . • • �c6 1 3 . . . itJd7 as played in some games is

best met with 14.h5! dxe5 ( 14 . . . itJxe5 transposes to the game) 1 5 .ic4! Probably strongest, as it forces Black into 1 5 . . . g5 16 .itJ d6 when I am very optimistic about White's position. The next moves will be ixfS, �d1 , or similar. White is winning. 1 3 . . . dxe5 14 .�xe5 f6 1 5 .VNc7 leads to a lost endgame for Black. 14.h5!

A wonderful move. White is not after material gains, but instead aims at the naked black king. Only at the absolutely correct moment will he take the exchange. Right now after 14.exd6 if6 Black is allowed unnecessary counterplay. This is the kind of game Black wants, why give it to him? 14 ••. �xe5 15.f4 .id7

This allows White to get rid of his only strange piece, the knight on b5 . But 15 . . . itJ d7 allows a winning attack. 16 .id3 itJ c5 ( 1 6 . . . if6 17.hxg6 ixc3t 18 . It>d1 fXg6 19 .ixg6+-) 17.hxg6 itJxd3t 1 8 .VNxd3 fXg6 19 .ixfS ixfS 20. itJxd6 and the endgame after 20 . . . ixd6 2 1 . VNxd6 VNxc3t 22. 1t>f2 �xc2t 23. lt>gl h5 24.VNd8t It>g7 25 .lt>h2 wins for White. The line analysed to the end goes like this : 25 . . . �f5 26. �hc1 �xf4t 27. lt>h1 It>h6 28 .� f1 VNe5 29. �ac1 and Black loses his bishop. 16.fx:e5 ,ixb5 17.,ixb5?!

17 .ixfS! is the superior move order, where Black is just lost.

17 • . . �xb5?! After this it is all over. 17 . . . �fc8!?

1 8 .exd6 �xc3 19 .id2 �xg3 20.ixa5 ± where White might have a few technical difficulties, was maybe the last hope!? IS..ixfB gxf8 19.exd6 .igS 20.gdl

Now White is simply winning. 20 •.. h6 21 .d7 �f5 22.gd3 gdS 23.hxg6 �e4t 24.�f2 f5 25.�f3 �a4 26.�e2 �f4t 27.:1�f3 �d6 2S.gdl �e7 29JUd3 It>g7 30.gel Elxd7 31 .gxd7 �xd7 32.�xe6 1-0

The final game of this chapter deals with the critical lines of the Pin-variation. I cannot promise White an easy advantage in these lines. They are still developing and Black players are learning more and more about the defensive possibilities in this line.

Game 56 Varga - Horvath Hungary 199 1

1 .e4 c5 2.�f3 e6 3.d4 cxd4 4.�xd4 �f6 5.�c3 .ib4!? 6.e5 �d5 7.,td2 �xc3 S.bxc3 ,te7

8 . . . ifS!? is a really interesting idea that has been popularised over the last few years. I think White is maybe a little better due to his lead in development, but it is not so clear. 9 .id3 d6 (On 9 . . . itJc6 then 10 . itJxc6!? makes good sense. 1 0 . . . dxc6 1 1 .0-0 �c7 12 . � e 1 ;!;


Lakdawala - Serpik, Costa Mesa 2003. This kind of position is hard to evaluate. Black is happy with his better pawn structure, while White enjoys more space and a lead in development. I think the position favours White, as the open b-file compensates for the double pawns, and the black king will be under some kind of harassment later in the game. However, the inconvenience for Black is likely to be minor.) 1 0 .exd6 (Famous is 1 O.'lWe2 lLld7 I 1 . lLlxe6 'lWb6 1 2. lLl c7t 1-0. Kasparov - West, telex 1 977, but 1 0 . . . dxeS would seriously test White's idea. Probably the position is just unclear.) 1 0 . . . ,ixd6 1 1 .0-0 ( 1 1 . 'lWhS! is my preferred move here. The position is not really so clear. Both players have their chances, but White seems to retain slight pressure.) 1 1 . . .0-0 ( 1 1 . . .eS!? should be considered when the critical line is probably 12 . 'IW e2 'lWe7 ( 1 2 . . . 0-0 1 3 .lLlbS! is unpleasant for Black, based on 13 . . . lLl c6 14 .,ixh7t *xh7 1 S .'lWd3t e4 16 . 'lWxd6 a6 17.'lWxd8 �xd8 1 8 . lLl c7 �b8 19 .,if4 when White has the advantage.) 1 3 .�fe l where White might have the advantage, based on 1 3 . . . exd4?! 14 .'lWhS ,ie6 I S . �xe6 'lWxe6 1 6. �el 'lWxel t 17 .,ixel ±) 12 . 'lWhS g6 13 . 'lWh6± Sargissian - Bursteinas, Tallinn 1 997.White has a slight edge after the opening. However, all the lines are too difficult to analyse conclusively based on just one game. We will stop here with the conclusion that White should not fear going for this line.

9.Y!Yg4 �f8!? This also has to be taken into

consideration, one opening book says. Actually it is quite a good move. lO.i.d3

lO . . . d6?! I do not like this approach. Black

should be careful about how he opens the position.

1O . . . hS!? 1 1 .'lWe4 dS 12.exd6 ,ixd6 13 .0-0 lLld7 led to unclear play in Zyla - Sulskis, Swidnica 1999. This might very well be Black's best option in the Pinvariation.

Very solid is 1O . . . lLlc6 but I still think that White is better. Black has no active play and the pawn weaknesses cannot really be attacked. I 1 . lLlxe6 dxe6 12.0-0 'lWe7 ( 12 . . . hS!? 13 .'lWf3 'lWdS 14.'lWe2 h4 was Diaz Luzuriaga, Buenos Aires 1999, when I S .h3 was the more normal move, when I think White should be a little better. There is no reason to fear weakening the g3-square, as


Black has no way to exploit it.) 13 .f4!? b6 14. Elael c5 Black is completely oblivious to White's plans: 1 5 .f5 is a nightmare for Black. White won quickly in Kottwitz -Hoen, St Ingbert 1 994. 1 l .f4 �d7

l 1 . . .ctJc6 12.0-0 d5 ( 12 . . . ctJxd4 13 .cxd4 f5 14.Wlf3:t Leone -Galli, corr. 1 969) 13 .f5!? White goes for it. An alternative was 13 . Elae1 preparing this advance. (Also strong seems to be 13 .Elab l . 13 . . . Wla5 14. ctJxc6 ic5t ( 14 . . . bxc6 1 5 .f5 exf5 16.ixf5 was also not pleasant.) 1 5 .ctJd4 ixd4t 16.cxd4 Wlxd2 17.f5 Wlh6 1B .Wlf3 b6 19 .ib5 Wlh4 20. Elf2 Wle7, but Black had had enough and resigned in Bresadola - Vibranovski, corr. 1 996.) 13 . . . exES?! (13 ... ctJxe5 was the logical move. My analysis goes like this: 14.'!We2 ctJxd3 15 .fxe6! ctJe5 (maybe better is 1 5 . . . if6 16.cxd3:t) 16. Wlxe5 if6 17. Wle2 ixe6 lB . ctJ xe6t fxe6 1 9. Wl xe6 and I like White. The main point is 19 . . . Wle7 2o.Wlxd5 EldB 2 1 .Wlg5!±) 14. ctJxf5 Wlb6t 1 5 .'it>h1 g6 16.ih6t 'it>eB 17. ctJ g7t 'it>dB 1 B .Wlf4 ie6 19.c4± Spiridonov - Poulin, e-mail, 1999. 12.0-0 �c5 13J�ael

I like this approach from the Hungarian grandmaster. His attack has a very natural feel to it. However also successful was 13 .f5!? dxe5 (An improvement would have been to remove the queen from the excellent square g4 with 13 . . . h5!? 14. Wl e2 dxe5 1 5 .Wlxe5 id6 16 .Wle3 e5 17. ctJ b5 ctJxd3 1B .ctJxd6 Wlxd6 19 .cxd3 when

White is only very slightly better.) 14.fxe6 if6? ( 14 . . . f6 15 .Wlh5 ixe6 16 .ctJxe6t ctJxe6 17.Wlxe5 Wld6 1 B .Wle2 would just have been slightly better for White) 1 5 . Elxf6! Wlxf6 16.ig5 and White won comfortably in Chorfi - Nadli, Ronde 1 995. 13 ... �xd3 14.cxd3 h5

This look like a weakening move, but the computer's choice is not reassuring either.

14 . . . Wla5 is met very strongly by 1 5 .f5! All the pieces are ready, so why not? 15 . . . dxe5 (White also wins after 1 5 . . . exf5 16 . ctJxf5 ixf5 17.Wlxf5 Wlxa2 1 B .c4) 16.fxe6 f6 17.Wlh5 g6 1B . Elxe5 Why exchange queens? 1 B . . . Wlxa2 19 .Wlh6t 'it>gB 2o.ig5! ifS 21 .e7 and White is winning. 15.tfa�/±

The question is not if White is better, but how much he is better. 15 ... g6 16.:i:lbl d5 17.J.e3 b6 18.c4 ib7

It is hard to see how Black should have played differently. Now White crashes through with a very thematic attack. 19.£5! gxf5 20.�xf5 exf5

20 . . . dxc4 2 1 .Wlxb7 exf5 22. Elxf5+- The f7 -square is very weak. Black will not be able to resist the attack. 21 .tfxf5 tfe8 22.tfg6!

A brilliant move. White wins by force. 22 ... dxc4 23.e6 f6 24Ji:xf6t ixf6 25.tfxf6t �g8 26.id4 1-0

248 The Nimzowitsch Variation

The Nimzowitsch Variation

- By Jacob Aagaard

The Nimzowitsch Variation is a very provocative line, in some ways not dissimilar to the Alekhine Defence. Black allows White to advance his e-pawn with gain of tempo, hoping that it will be weak rather than strong on e5. When I started to work on this chapter I believed that this was a minor line and could be dealt with quickly. Instead it proved at times more difficult to approach than the feared Sveshnikov.

Against the Nimzowitsch variation l .e4 c5 2 . ltJf3 ltJf6!? I have decided to recommend the less principled 3 . ltJ c3! based on lengthy analysis and my own personal preference. I know that this move gives Black the chance to escape back to the main lines, but since we were ready to

play them just a move ago, we still should be now. Obviously it would be desirable to punish Black for his provocation, but if 3.e5 ltJd5 4. ltJc3 then 4 . . . ltJxc3! . This is, of course, what Black wants to do. The e-pawn has advanced too far, and White's pawn structure is damaged. White gets so-called "free play" , but I seriously doubt that this includes an advantage. 5 .dxc3 ltJc6.

This posltion has not give White a particularly good score. I generally do not like the position for White; I would much rather play the Black side, as in that case I would mind less if the game suddenly fizzled out into a draw! (4 . . . e6 5 .ltJxd5 exd5 6.d4 ltJc6 7.dxc5 ixc5 8 .�xd5 is notoriously bad for Black, but besides the Israeli 1M Mek who has Yz/4 with this system, no strong player seems to favour it regularly. Still I have included two games with it; game 59 and 60) . 4.d4!? cxd4 5 .�xd4 promises White a slight edge in


ECO, but maybe that is the only place. This method of play seems to be logical and gives White a good score, but I did not find it dangerous enough to Black.

Still, after 3 .lLlc3 there are ways for Black to continue with deviations from the main lines. And actually many Black players would choose to do so: with 3 . . . e6, inviting 4.e5 (which we accept) , with 3 . . . d5 !? , which is considered in our first game, and with 3 . . . lLlc6 4.d4 (4.ib5 is another popular move here, but I want to stay true to our repertoire) 4 . . . d5!?, which is underrated by theory. However, I still think White is a little better - see game 58 .

We will start by having a look at the counterintuitive 3 . . . d5!?

Game S7 Borriss - Bartsch Germany 2003

1.e4 cS 2.lLlf3 lLlf6!? 3.lLlc3! dS!? This is pushing it! Black decides to open

the position after a move like . . . c5, and obviously being behind in development. It is no surprise that aggressive play seems to give White an advantage.

3 . . . d6 4.d4 leads to the standard positions of the Sicilian. 4.exdS lLlxdS s.AbSt Ad7 6.lLleS!

I like this move and find it logically correct, but 6. Vff e2 is a possible alternative to this aggressive approach.

6 . . . hbS?! The alternative is 6 .. . lLlf6 and now

White has two options, where only the most aggressive seems to work.

a) 7. lLlxd7 lLlbxd7 8.d4 cxd4 (8 . . . a6 9 .ixd7t (9.ie2!?) 9 . . . lLlxd7 1 O .d5 g6 1 1 .ig5 Ag7 12 .0-0 f6?! ( 1 2 . . . 0-0 1 3 .d6;!;) 13 .if4 0-0 14 .Vffe2± Poulsen - Bach, Esbjerg 2004) 9 .Vffxd4 a6! (9 . . . e6 10 .ig5 a6 1 1 .ixf6 gxf6 12 .ixd7t Vffxd7 13 .Vffxf6 �g8 14 .0-0 ( 14 .Vfff3!?) 14 . . . Vffc6 1 5 .g3 ig7 16 .Vffh4 Axc3 17.bxc3 Vffxc3 1 8 .Vffxh7 �h8 19 .'IMle4± Papin - Grigorov, Chalkidiki 2003.) 1 0 .ie2 ( 10 .ixd7t 'IMlxd7 1 1 .ie3 'IMlc6= Bielczyk - Kunin, Schwarzach 2002.) 10 . . . 'IMlc7 1 1 .0-0 e6= Almagro Uanas Solaesa Navalpotro, Madrid 2003.

b) 7.Vfff3! 'IMlc7 8 .0-0 (8 . lLlxd7 lLlbxd7 9.d3 (9 . . . a6 1 0.ixd7t Vffxd7 1 1 .ig5 was unpleasant for Black in Gentinetta - Perez, e-mail 2000. 9 . . . 0-0-0 1O .if4! e5 1 1 .ig3 a6 12 .ic4 lLlb6 1 3 .0-0-0

250 The Nimwwitsch Variation

id6 14 .ii.h4± Shirov - Fernandes, Elista 1 995.) e6 1 0.ii.f4 '!Mfcs l 1 . ltJ e4t Ganguly - Sriram, Calicut 2003.) S . . . e6 (S . . . '!Mfxe5? 9. '!Mfxb7 +-. I see no possible compensation. Remember 9 . . . ltJ g4? 1O .'!MfcS mate.) 9J'!e l ! In this position it is very difficult to find a way for Black to escape from the opening without permanent scars in his pawn structure or loss of material.

a) 9 . . . ii.e7 1 0.a4? was played in Gusev Krivobokov, Tula 2002. Instead 10 .'!Mfg3! would have won a pawn for very little. 1 0 . . . ii.xb5 ( 1 0 . . . ltJc6 1 1 . VNxg7 l:"!f8 12 . ltJxd7 VNxd7 1 3 .ltJe4± . 1 0 . . . 0-0 l 1 . ltJ xd7+-) l 1 . ltJxb5 '!Mfb6 12 .VNxg7 l:"!f8 13 .a4± .

b ) 9 . . . ltJ c6 10 .ltJxd7 ltJxd7 l 1 .ixc6t with a very nice endgame.

c) 9 . . . ii.d6 does not work because of 10 .ii.xd7t! ltJ bxd7 l 1 . ltJ b5 VNbS 12 .ltJxd6t VNxd6 13 . VNxb7 0-0 14 . ltJxd7 ltJxd7 1 5 .d3+-. It is just a pawn.

d) 9 . . . ii.xb5 10 .ltJxb5 '!Mfb6 l 1 .a4! gives White a position that is dose to, or perhaps even directly, winning. It is all based on trapping the black queen in an unusual manner. ( 1 1 . ltJ a3 ltJ bd7 12 . ltJxd7 ltJxd7 13 . ltJ c4 VNa6 14.b3 ie7 1 5 .ib2 0-0=) l 1 . . .ltJc6! Only move. ( 1 1 . . .ii.e7 12. ltJ c4± . 1 1 . . .a6 is met strongly with 12 .a5 ! ! '!Mfxb5 1 3 .c4 '!Mfb4 14.b3+-. The next move is 1 5 . l:"!a4.) 12 .ltJxc6 ( 1 2. ltJ c4 '!MfdS 13 .'!Mfg3 Black's position is obviously problematic, still it is not entirely dear. 1 3 . . . ltJd4! 14. ltJ c7t rJJe7 1 5 . ltJ e3 l:"!cS 1 6. ltJ b5 rJJ eSt) 12 . . . bxc6 13 . ltJ a3± .

e) 9 . . . a6! This i s the best try. Actually the only one I cannot refute directly. 1 0. ltJ d5!? Otherwise White has nothing. ( 1 0.ii.xd7t ltJbxd7 l 1 . ltJ c4 ltJ b6!=) 10 . . . ltJxd5! (Black is forced to sacrifice l:"!hS. 10 . . . '!MfdS l 1 . ltJxf6t gxf6 12 .ltJxd7 ltJxd7 13 .ii.xd7t VNxd7 14 . VNxf6 l:"!gS 1 5 . VN f3 ± offers no compensation.) 1 1 .ii.xd7t ( 1 1 . VNxf7t rJJdS 12 .ixd7 ltJxd7 13 .VNxe6 ltJxe5 14 . l:"!xe5 ltJ f6 1 5 .d4 '!Mfd7!'t') l 1 . . . ltJxd7 12 .ltJxf7 ltJ7f6 ( 12 . . . l:"!gS 13 . l:"!xe6t ltJ e7 14.d4! with a winning attack) 1 3 .ltJxhS 0-0-0 14. VNh3 ltJ f4 1 5 . VNh4 This position is, of course, hard to evaluate. If White is afraid of going for this kind of adventure, then S .ltJxd7 offers a slight edge without any complications. However I truly believe that White is better here. 1 5 . . . id6 16 .d4 l:"!xhS 17.dxc5 ii.e5 I S .ii.xf4 ii.xf4 19 . 1:"!xe6± is just one line where White


enjoys the strength of a rook vs. two minor pieces which have no good squares. 7.ti'S!

This is a famous trick-shot. 7 • • • f6 8.c!LJxb5 c!LJa6

This is the gambit choice behind 6 . . . ixb5, but it is flawed. Unfortunately for Black it seems that the best move is 8 . . . fxe5 9 .ti'xd5 ti'xd5 10. c!LJ c7t md7 I l . lLlxd5 lLlc6 1 2.d3 e6 1 3 .lLle3±/± and the endgame is great for White. Laznicka - Malmqvist, Marianske Lazne 2003.

8 .. . lLlb4 9 .�xb7! just wins for White. 9 . . . fxe5 1 0. lLl c7t mfl 1 1 .ti'f3t mg8 12.�b3t lLld5 1 3 .lLlxa8 lLlc6 14 . lLlc7 lLld4 1 5 .�xd5t �xd5 16 .lLlxd5 lLlxc2t 17. mdl lLlxaI 1 8 .b3+-.

9.ti'h5t! Otherwise nothing makes a lot of sense

to me. Other moves have been played, but they are not worth our time. 9 ••• g6 10.c!LJxg6 hxg6 1 l .ti'xh8

Ka-ching!

1 l • • • ti'd7 12.c!LJc3 �e6t 13.mfl c!LJdb4 1 3 . . . lLl ab4 14 .�h3!? ( 14. lLl xd5 ti'xd5

1 5 .ti'h4 md8! !+ . 14 .�h7! however looks strong. Black has no simple response as after 14 . . . 0-0-0 1 5 . ti'xg6 the queen is back in the game.) 14 . . . ti'a6t ( 14 . . . ti'xh3 1 5 .gxh3 lLlc7 is in the spirit of the position, but White should be better here as well.) 1 5 .d3 �d8 16 .ti'g4 f5 17 .�e2+-, Carlsen - Runde, Norway 2002. 14.ti'h4!

This is the way to play. The queen enters the game and slows down Black's initiative.

Less clear is 14 .�h3 ti'xh3 1 5 .gxh3 lLlxc2 16 .�bl lLl ab4 17 . �gl mfl� Dirr - Bartsch, Germany 2003. The pawn structure is a true nightmare, and White cannot develop without compromising it further.

14.d3! ? looks sound. However, after 14 . . . lLlxc2 1 5 .ih6 O-O-O! Black certainly has a lot of play. One plausible line is 1 6. �c l ixh6 17.�xh6 lLlab4 18 .�xg6 lLlxd3 19 .�b l lLl f4 20.�e4 �a6t 2 1 . mgl lLlh3t! 22.gxh3 �g8t 23.�g4t �xg4t 24.hxg4 ti'd3t' . 16 .�dl looks safer, but still Black obviously has play. 14 • • • 0-0-0

14 . . . lLlxc2?? 1 5 .ti'a4t is an important, though simple, point. 15.ti'e4!

Offering the exchange of queens without ruining the pawn structure. 15 • • • ti'd7

It is easy to criticise this, but Black's position is bad. 1 5 .. .'� xe4 1 6. ltJ xe4 ltJxc2 17 . Eibl ltJcb4 1 8 .a3 ltJd3 1 9.h4± . Here I do not trust Black's compensation. A queen is obviously needed, or a general target. Both ltJa6 and i.fB are out of play. 16.a3 f5 17.'l«e4 �xc2 18.l3:hU/±

White has escaped from the opening an exchange up. It really is that simple. If the advantage is t or ± is not so clear, but does it really matter?

18 ••• <.i?h8 19.d3 e5 20.�g5 'i'xd3t 21 .'l«xd3 l3:xd3 22.M! �xa3 23.hxa3 l3:xc3 24.h5 gxb5 25.l3:xh5 i.d6?

Losing a piece. But afrer 25 . . . <.i?c7 26. Eih7t <.i?c6 27J �hxb7 Eixa3 28. Eixa7 c4 29. Ei£7 i.d6 30. EixfS White is a likely winner. 26.l3:dl! 1-0

I clearly have my doubts about 3 . . . d5, though an outright refutation probably cannot be found. However the next line was a real nightmare to prepare for. It took me two days to decide that the uncommon 6.dxc5!? is White's best try for an advantage.

Game 58 Ismagamhetov - Palit Kuala Lumpur (U- 1 6 Wch) 2002

1 .e4 e5 2.�a �f6 3.�c3 �c6 4.d4 d5!? The more I studied this move the more surprised I was. It is actually not stupid at all! 5.exd5 �xd5 By transposition we have reached the position afrer 3 . tlk3 ltJc6 4.d4 d5 5 .exd5 ltJxd5 .

6.me5!? This is the best attempt for an advantage

Experts vs. the Sicilian 2S3

I can offer here. Basically I think Black should be able to scrape a draw quite often, but still it is a sad life . . .

6 .ibS? ! ig4! leads to no advantage for White, only problems. I wonder why ig4 has been played only a few times.

6. lLlxdS �xdS 7.ie3 cxd4 8 .lLlxd4 a6 9 .ie2 eS also seems reasonable to me. 6 ... tt)xc3 7.'i'xd8t tt)xd8!

7 .. .'itt xd8 is the main alternative. My feeling is that White should gain a small plus in all lines. 8 .bxc3 f6 (8 . . . eS 9 .lLlgS li7e8 10 .ic4 lLld8 1 l .f4 h6 12. lLl e4± Santo Roman - Hausner, Prague 2000.) 9 .ie3 eS 1 0.ic4 ( I o. lLl d2!? is a very natural alternative with the score of 3/3. 1O .. . ie6 l 1 .ic4 ixc4 12 . lLlxc4 li7c7 13 .0-0-0 ie7 14J'1d3 gad8 I S . ghdl ± Feygin - Mek, Belgium 2004.) 1 0 . . . li7c7 ( I o . . . ig4 as in Collins - Schalkwijk, Corr. 200 1 is best met with l 1 . lLld2 li7c7 12.f3 ifS 1 3 .lLle4± where Black soon could be much worse) l 1 . lLld2 ifS 12.0-0-0 ie7 1 3 .f3 gad8 14 .h4 ic8 I S . lLl e4 h6 16.g4 hS 17. gxd8 lLlxd8 1 8 .gS± Rowson - Murey, Pula 2002. 8.bxc3 f6!

Building a big centre is the best plan for Black.

On 8 . . . g6 then 9 .ibSt! seems to be the best way forward. 9 . . . lLlc6 (9 . . . id7 l O.ixd7t Iitxd7 l 1 .ie3 gc8 1 2.0-O-ot li7e8 13 . gd3±/±) 1 0. lLl eS ig7 l 1 . lLl xc6 a6 12 .ia4 id7 13 .gb l ixc3t 14.id2 ixd2t I S . li7xd2 bxc6 16 . gb7± Froeyman - Hajenius, Belgium 200 1 .

8 . . . id7 9 .ie3! is known as better for White. The following lines shows why:

a) 9 . . . g6 1 0.0-0-0 ig7 l 1 .id4 ih6t 12 . li7b2 0-0 1 3 . lLl eS ie6 14 .ibS f6 I S . lLl d3 idS 1 6.f3 eS 17.if2± Krnic Wedberg, Eksjo 1 978.

b) 9 . . . f6 1 0 .lLld4 gc8 l 1 . lLlb3 eS 12 .ic4 b6 13 .ia6 gb8 14.0-0± Bravo Barranco - Paredes, Barcelona 1 996.

c) 9 . . . gc8 1 0. lLl eS ia4 l 1 . gbl a6 ( I l . . .e6 12 . gb4 ixc2 13 . li7d2 ixcs 14 .ixcS gxcS I S . gbS !± MUller -Bastian, Germany 1 99 1 ) 12 .id3 e6 13 . gb4± ic6 14.0-0 ie7 I S . gfb l 0-0 16 .lLlxc6 gxc6 17 .ie4 Ek7 1 8 .ixb7 ixcs 19 .ixcS gxcS 20.ixa6+- Romero Holmes - Casafus, Dubai (ol) 1 986. 9.ie3 e5

lO.tt)d2 The most flexible. Also played has been

10.0-O-0?! ' With this White loses the option of invading down the b-file, and with that all chances of an advantage.


1 O . . . .ld7 ( l0 . . . lLle6 l U',!d5 lLl c7?! 12J"!d2 lLl e6 1 3 . .lb5t �f7 14 . .lc4 �e8 1 5 . ghdl ± is not what Black is looking for. Those "repetitions" make his position worse continually.) l 1 . lLld2 gc8? 12. lLl e4 lLl e6 ( l2 . . . f5! ? was a better try. After 13 . lLld6t .lxd6 14.cxd6 b6 1 5 . �b2;!; White is better all the same.) 13 . gd5 .lc6 14. lLl d6t .lxd6 1 5 . gxd6;!; Vallejo Pons - Mek, France 2003. However 1 1 . . .b6! would have kept Black at least equal. The main point is 1 2. lLl e4 lLl b 7! with compensation. I have analysed this line somewhat and equality was the best I found for White, and that was not so evident in all lines! 10 ••• .100

10 . . . lLl e6 l 1 . lLl e4 .ld7 12 .gbl .lc6 as in Schramm - Pawlitzki, Germany 1 992. White might have a slight advantage, but I am not even certain. However all moves are open to discussion. 1 l .J.b5t J.d7

12.hd7t?! White has the chance to play for

an advantage only through his lead in development. Therefore the right choice was a developing move, and not a move that develops the opponent's king.

1 2 .gb l ! would guarantee White a slight advantage: 1 2 . . . gc8! ( l 2 . . . .lxb5 1 3 . gxb5 �d7 14 . �e2 and 1 5 . ghb l i s clearly good for White.) 13 .i.xd7t �xd7 14. lLl b3;!; Now there is no a7-a5, so I think that White has the better prospects. One line could be 14 . . . lLl e6 1 5 . �e2 i.xc5 1 6 .lLlxc5t lLlxc5 17 . .lxc5 gxc5 1 8 . gxb7t �e6 19 . 9xa7 gxc3 20. gcl ;!; . 12 ... �xd7 13.0-0?!

1 3 . �e2 is also better here. 13 ... lLlOO 14JUdl �c6 15.lLlb3 .1e7?? Black realises that he cannot allow White to invade to the 7th rank. However, ifhe had prepared it, taking the c-pawn would have been fine. 1 5 . . . a5! 1 6.a4 J.xc5 1 7. lLl xc5 lLlxc5 1 8 .J.xc5 �xc5 19 . 9d7 b5� and the distant passed pawn and active king are quite good for creating counterplay. 16.lLla5t �c7 17.lLlxb7!± And White won.

In the next two games we shall see the heavily theoretical main line of the Nimzowitsch Sicilian. I actually think that the previous line represents Black's best option here, and that the next two games are fought over in somewhat dubious territory.


Game 59 Braun - Choroba e-mail 2002

1 .e4 c5 2.ctla ctlf6 3.e5 ctld5 4.ctlc3 e6 Our move order is of course with 3 . ctl c3 e6 4.e5! ttJd5 , when we can continue with the game.

However in our move order it would also be possible to play 4.d4 aiming for a normal Sicilian. If Black is a true Nimzowitsch-Sicilian player he can try 4 . . . d5!? I now prefer 5 .exd5 , but let us look at both lines:

a) 5 .e5!? is possible, but now we are in the French Defence. Positions such as 5 . . . ctl e4 (5 . . . ctl fd7 6.dxc5 ixc5 7.id3 ttJ c6 B .if4 f6 (B . . . 'Wb6?! 9 .0-0 'Wxb2 is a bad idea. White wins with 10. ttJ b5 0-0 ( l 0 . . . @f8 l 1 .id2! d4 12 .'We2 and the queen is trapped.) l 1 .ixh7t! @xh7 12.'Wd3t g6 1 3 . �fb l ttJb4 14 . ttJg5t @g7 1 5 .'Wh3+-) 9.exf6 ttJxf6 10.0-0 0-o l 1 . ttJe5 Black should be OK here, still the position is not that easy. 1 1 . . . ttJ e4? 12 .ixe4! �xf4 1 3 .ttJd3, White is slightly better in Zatonskih - Tessier Desrosiers, Kapuskasing 2004.) 6.id3 ttJxc3 7.bxc3 c4 B .ie2 ie7 9 .h4 h6 10.h5 ttJ c6 1 1 . �h3 Wfa5 12 .id2 id7 13 . �g3 if8 14. @f1 0-0-0 1 5 . @gU Baklan -Danneel, Ghent 2003. It is all a matter of what kind of position you enjoy playing, of course.

b) 5 . . . ttJxd5 6. ttJxd5 Wfxd5 (6 . . . exd5 7.ib5t id7 B .ie2!? Hardly the

only way to play. B . . . ie6 9.0-0 ie7 1 0.dxc5 ixc5 1 1 . ttJ g5 ttJc6 12 .ttJxe6 fxe6 13 .ig4 'Wf6 14 .'We2 ttJd4 1 5 .'Wd3 0-0 16 .ie3 �acB 1 7 .�ae 1 , with a slight advantage for the first player, Filippov -Kohanchik, Moscow 2002.) 7.ie3 cxd4 B .ttJxd4± a6 9 .ie2 e5? ! . But it was not so easy for Black anyway. 1 O.if3 Wfa5t l 1 .id2 ib4 1 2.0-0 ixd2 1 3 .ttJb3 Wfc7 14 .'Wxd2 0-0 1 5 . � fe l ie6 16.Wfa5 WfcB 1 7.'Wxe5, White is objectively winning, Tiviakov - Mek, Vlissingen 2003. That White did not win this game is not easy to understand. 5.ctlxd5 ad5 6.d4 ctlc6 6 . . . d6 has long been held in disregard because of

7.ib5t! giving White a slight but lasting edge in all lines. 7 . . . ttJc6 (7 . . . id7 B.ixd7t 'Wxd7 9.0-0 ttJc6 1 0.exd6 ixd6 1 1 . �e 1 t ttJ e7 1 2.dxc5 ixc5 1 3 .ie3± with a sad endgame for Black in Nemec - Volf, Czech Republic 2003.)


S.o-o !i.e7 9.c4 and we now have: a) 9 . . . !i.e6 10 .te3 llNb6 l 1 .a4 a6

1 2.a5 Wic7 1 3 .exd6 llNxd6 14.dxc5 WidS 1 5 .!i.xc6t bxc6 1 6. ltl e5 llNc7 1 7.Wia4 :1'!:cS I S .cxd5 !i.xd5 1 9. 1tl c4, White is better, Unzicker - Pomar Salamanca, Bad Aibling 1 965.

b) 9 . . . dxc4!? is possible. White however retains the better play after 1 0.d5! a6 1 1 .!i.a4 b5 12.dxc6 bxa4 1 3 .exd6 Wixd6 14 .Wixa4 0-0 1 5J:idl llNc7 16 .!i.g5 ;!;) Old theory says:

c) 9 . . . a6 10 .!i.xc6t bxc6 l 1 .cxd5 cxd5 1 2.exd6 Wixd6 and now White has two ways to go.

d) 13 .!i.g5! ? txg5 14 .ltlxg5 0-0 1 5 .dxc5 Wixc5 1 6.llNd3;!; .

c2) 1 3.dxc5 llNxc5 14 .!i.e3 llNb5 1 5 . ltl d4 with a clear edge for White according to Boleslavsky. However I think that evaluation is a bit optimistic. A minor advantage for White is probably more accurate. One game went 1 5 . . . Wixb2! ? 1 6. ltl c6 !i.e6 (16 . . . !i.f6 17.llNxd5 te6 I s .Wid6 Wib7 1 9J:ifd, gives that minor ;!; I'm talking about.) 17. ltl xe7 rtJxe7 I S .!i.c5t rtJd7 1 9.!i.d4 Wia3 20J:iel � ghgS 2 1 . llNb l gacS 22.te5 a5 23.Wib7t rtJdS 24. gab l llNc5 25. gb6 Wie7 26.Wia6 a4 27.!i.d6 Wid7 2S.!i.b4 WieS 29. gd6t 1-0. Herbst - Rossell, e-mail 2000. 7.dxc5 hc5 8.Wixd5 'i'b6 S . . . d6 is maybe slightly sounder, and portrayed in the next game. 9.J.c4 hflt lO.rtJe2 0-0 l um . .lc5 l2.�g5

12 • • • ltld4t This main line move does not seem to

lead anywhere. I have a bad feeling about Black's position in these lines.

12 . . . ltlxe5!? is a very tricky move. According to theory it is just winning for White. However, the theoretical line leads to unclear play or a forced draw, if Black finds some simple improvements. 13 .Wixe5 d5 14 .txd5 ( 14.Wixd5 tg4t 1 5 . gf3 gadS! was unclear in two mid SOs correspondence games from Prieto Fernandez. However I am not sure that White cannot find an advantage here as well, though it would be under very unclear circumstances. In Beating the Sicilian 3 Nunn and Gallagher only give a game with 1 5 . . . !i.g 1 , which to me seems less relevant. To be fair, I must point out that the databases over the last few years have improved immensely.) 14 . . . tg4t 1 5 . rtJd3 was successful in Kaidanov -Kreitner, Chicago 1 995 , which made


Kaidanov believe it was close to winning. However ( 1 5 . We1 ! ? E:ae8 16.i.xf7t Wh8 17.'�'xe8! ( 17.i.xe8?? 'lMra5t! !-+) 17 . . . E:xe8t 1 8 .i.xe8 h6 19 .h3 ( 1 9. tLl e4 i.e3co) 1 9 . . . hxg5 20.hxg4 'lMre6t=) 1 5 ... E:fe8!! gives White only perpetual check with 1 6. tLl xf7, as 1 6.i.xf7t Wh8 17.i.xe8?? E:d8t! is mate in a very few moves.

1 3 .E:f4!? might be the right move.

Now we have: a) 1 3 . . . tLlxc4 14 . E:xf7+-. b) 13 . . . 'lMrd6 14.'lMre4 g6 1 5 .tLlxf7 tLlxf7

16. E:xf7 E:xf7 17.'lMre8t 'lMrfB 18 .i.xf7t Wg7 1 9. 'lMrxf8t i.xf8 20.i.d5± . This endgame is terrible for Black: behind in development and with a clearly inferior pawn structure.

c) 1 3 . . . d6 14 .'lMre4 g6 1 5 .tLlxf7! i.g4t 16Jhg4 tLlxg4 ( 1 6 . . . tLlxc4 17. E:xg6t hxg6 1 8 . 'lMrxg6 mate.) 1 7. tLl g5t Wh8 18 .'lMrxg4 E:ae8t 19 . Wd3 i.e3 20. tLl e6 Having used quite some time on this

position, I think it is clearly better for White. The main line goes 20 . . . d5 (20 . . . E:f4 2 1 . i.xe3 'lMrxe3t 22. Wxe3 E:xg4 23.i.d5 ±) 2 1 . i.xd5 E:xe6 22.i.xe6 E:d8t 23.i.d7! i.xc1 24. E:xc1 'lMrd6t 25 .'lMrd4t 'lMrxd4t 26. Wxd4± and White has great chances to win the endgame.

d) 1 3 . . . h6 Probably the best option. 14. tLlxf7 ( 14 . 'lMrxe5?! d5!Always thismove. 1 5 .i.xd5 hxg5 1 6. E:xf7 E:xf7 17.'lMre8t Wh7 1 8 . 'lMrxf7 i.g4t�. I have my doubts about White's chances of surviving such an attack.) 14 . . . tLlxf7 1 5 . E:xf7 'lMre6t Only move, which is actually not such bad news for White. 1 6. 'lMrxe6 dxe6 17. E:xf8t Wxf8;!;. Here White can play many things. The endgame is slightly preferable, but nothing fantastic. It is possible to study the lines further, but my general feeling is that this is the kind of thing White will be able to get. A complete bust is hard to find. 13.Wdl tLle6 14.�e4 d6

14 . . . i.g1 has only been played once. In that game Black even managed to get the advantage, but logical play wins for White. 1 5 . tLl d6! An excellent square for the knight. 1 5 . . . i.xh2 ( 1 5 . . . 'lMrd4t? 1 6. 'lMrxd4 i.xd4 17 .E:e1 ± and Black cannot easily get his pieces into play. ) . Now 16.g3 'lMrc6 17.i.f4 was played in Kopelevich - Muehlenweg, e-mail 2002. 17 . . . 'lMrxd5t 1 8 .i.xd5 g5 ! ?co is not what White is looking for. 1 6.i.d3 also looks tempting, but I could not make it work. 1 6 . . . 'lMrc6! 17 .'lMrxc6 ( 17.c4 h6 1 8 .tLlf5

258 The Nimwwitsch Variation

E:e8 1 9 .i.xh6 �xd5 20.cxd5 i.xe5!+) 1 7 . . . dxc6 1 8 .E:h1 i.xe5 1 9 .i.xh7t Wh8 20.i.g6t Wg8 2 1 .i.xf7t A wonderful combination, but only a draw. 2 1 . . . E:xf7 22. E:h8t Wxh8 23. tLlxf7t Wg8 24. tLlxe5=.

But 1 6. E:h ! ! , simple and strong, seems to be too dangerous for Black.

My analysis gave these lines: a) 16 . . . ,ig3 17.,id3! h6 (17 . . . f5

1 8 .tLlxf5) 18 .tLlf5 ,if4 19 . tLl e7t Wh8 20.�e4+-.

b) 16 . . . h6 loses in similar ways to the other lines. 17.,id3 ,if4 18 .�e4 f5 19.exf6 �xd6 20.fxg7 tLlxg7 2 1 . �h7t Wf7 22. E:f1 and White wins material.

c) 16 . . . �f2!? 17.,id3 ( 17.�e4 f500) 17 . . . tLlf4 1 8.,ixh7t! !Averynice refutation. ( 18 .,ixf4 �xf4=) 18 . . . Wxh7 19 .�d2 �xd2t 20.,ixd2. It is obvious that White has a fantastic position once he regains the material, but he also has a fantastic attack after 20 . . . tLle6 2 1 . E:xh2t Wg6 22.a4! with

E:a1-a3-g3 coming, deciding the game in White's favour.

d) 16 . . . h5!? appears to be the hardest move to kill. The point is that the pawn is better placed on h5 than on h6 in many lines. However I seem to have found a way to grind Black down. 17.a4! A whole new resource. ( 17 .,id3 g6 surprisingly seems to lead nowhere: Black's position is hard to crack. So what I thought was that White should get E:a1 into play, since this was the problem with all my previous attempts. And I should know, having lectured continuously on the importance of this simple rule in my books.) 17 . . . .tf4 18 . E:a3. I think White must have the better chances here. Moves like E:h3 and E:B are coming, and f7 seems bound to fall. My two main lines are:

dI ) 1 8 . . . �c6 19 .�xc6! dxc6 20.!xe6 fxe6 2 1 .,ixf4 E:xf4 22. E:xh5+-. White clearly wins positionally, but he has a winning attack as well.

d2) 1 8 . . . g6 19 .E:B ,ixc1 20. Wxc1 and White has an absolutely winning position. The next few moves are likely to be � d3 and !xe6. One line is 20 . . . �b4 2 1 . � e4! with the idea of various sacrifices: 2 l . . . Wg7 22. tLlxf7 E:xf7 23. E:xf7t Wxf7 24.!xe6t +-. 15.exd6 1:!d8 16.1d3 .bI6 17.YlYh5 f5 18.ttJxd6 YlYxd6 19.YlYxf5 YlYxh2

I do not believe that Black's position can be saved anymore.

Another try has been 19 . . . tLlf8, but 20.�f7t Wh8 2 1 .�f4 seems to put Black


a pawn behind. It is really as simple as that. Here are a few lines to prove it:

a) 21 . . .�g6 22.�xd6 ig4t 23.lt>e1 Eixd6 24.if4 Eif6 25.ig3+- Zanetti -Corinthios, corr. 1 9B6.

b) 2 1 . . .�xf4 22.ixf4 if5 Repp -Boeckler, corr. 1 993. Now strongest is 23. It>e2! EieBt 24. 1t>f2+- with a winning endgame.

c) 2 1 . . .�e7 22.id2!. White is best off developing. 22 . . . �g6 23.'�·g5 �xg5 24.ixg5 Eid5 25.ie3 ig4t 26. It>cl �e5 27.ie4+- Wolff - Izumikawa, USA 19B7.

d) 2 1 . . .�c5 22.ie3 �h5t 23. lt>e1 ! . Simplest. There i s security to be found on the kingside. 23 . . . id7 24. 1t>f2 �g6 25.�g5 EifSt 26. lt>gl +- Lehner -Kummer, Hartberg 1992. 20.%Yf7t It>h8 21 .ig5 l:'!g8 22.ie3 �xg2

The only move. The alternative 22 . . . � dB 23.�f4 �xf4 24.Eixf4± has given White a clearly better endgame in many games. There is no reason to know more than this.

23.�h5! The old move. An impressive game

was 23. lt>cl ! ? �d5 24. Eif5 �h1t 25 . 1t>d2 �h2t 26. Eif2! %Yh4 27. Eig1 g6 2B. Eih1 !! The final blow. 2B . . . %Yb4t, but Black resigned because of 29. 1t>c1 Eig7 30. Eixh7t! Haba - Kummer, Austria 1 99B.

However, there is still life in Black's position, despite the computer's disbelieE 23 . . . %Yh2! 24.�f3 %Ye5 25 . Eih 1 EifS! ! was an impressive discovery. I am not sure White is better after this. 26. Eixh7t It>gB 27.�h3 � d4 2B. EihBt 1t>f7 29. EixfSt 1t>xf8 30.%Yfl t �f5 3 1 .if4 �d5 32.ic4 %Ye4 33.id6t It>eB 34.ih2 ie6 35 .id3 Y2-V2. Elburg - Turati, e-mail 2000. 23 ... g6 24 . .td4t �g7 25.hg7t It>xg7 26.�e5t It>h6 27.%Ye3t!

27. �f4t It>g7 2B. %Yf6t It>h6 29. %Yh4t It>g7 30.%Yd4t It>h6 leads nowhere. 27 ... lt>g7

27 . . . g5 2B. lt>cl ! (2B . Eif6t Eig6 29. Eif2 �h3 30.%Yxh3t ixh3oo) 2B . . . ih3 29.�e7 E:g6 30. Eid1 ± is similar to the game. Black is suffering and there is no end in sight. 28.lt>cl!

This quiet waiting move underlines the real problem in Black's position, which is not the open king, but that he is unable to develop his pieces sensibly. 28 ... ,tf5

This must surely have hurt, but Black cannot save the position. 2B . . . ih3 29. Eif2 �glt 30. lt>d2 %Yg4 3 1 . Eih1 �b4t


32 .';iJdl ig4t 33. mcl h5 34."W'e5t mh6 35 .E1f7 and the king is toasted.

28 . . . "W'h2 29."W'd4t mh6 looks like a defence, but White wins elegantly with 30.'�d5! ! ih3 3 1 . E1hl "W'g3 32.ifl +-. 29.§'d4t mh6 30.E1g1 "W'h3 31 .hf5 gxf5 32.§'d2t £4 33.§'xf4t mh5 34.§'e5t mh6 35.E1e1 !

White has succeeded in stripping Black's king completely, and material is still level. If White succeeds in getting E1al into play he will win very easily. This is exactly what happens in the game. There are many lines possible in the coming moves, all leading to White's satisfaction. I have chosen not to include them, as this position is clear enough to make an evaluation on, and as almost any move is possible all the time. The game played by White here is a great achievement, especially as both players are rated around 2000. However, being an e-mail game I cannot help wondering

if they had some help from the silicon monsters . Advanced chess does seem to live quite well in e-mail tournaments. From a theoretical point of view it is, of course, a great thing that this game exists. 35 ••• gac8 36.mbl gc6 37.a4 ggS 38.§'e7 §'f5 39.ga3 §'xc2t 40.mal §'g2 41 .§'mt mh5

4 1 . . . E1g7 42.13:ae3 E1a6 43. §' f4 t E1g5 44. E1e7 and Black's position is collapsing. 42.ga 1-0

In the last game of this chapter we will have a look at the double pawn sacrifice line with 8 . . . d6. This is not played as often as 8 . . . "W'b6, probably because it is less dramatic. White should be able to prove an advantage by keeping one of the pawns and torturing Black in endless endgames, until Black players stop playing this line.

Game 60 Hlavac - Bazant Czech Republic 2000

1 .e4 c5 2.�a �f6 3.e5 �d5 4.�c3 e6 5.�xd5 exd5 6.d4 �c6 7.dxc5 .bc5 8."W'xd5 d6!?

This move has a better theoretical reputation than 8 . . . "W'b6, which is probably justified. I still think that White should get the advantage. 9.exd6 §'b6 10."W'e4t


This is the main response, and probably also the best. 10 . . . .i.e6 l 1 .lWh4

1 1 . . . £6 Black has many alternatives, from

which the first is the best, and the last is the worst possible move in the position.

On l 1 . . ..i.xd6 then 12 . .i.d3!? is a fine move. ( 1 2 . .i.e2 is played more often, but after 1 2 . . . if5! ? as in David - Luther, France 2003 it is not easy to prove an advantage, nor after the main line 1 2 . . . ie7 13 .lWe4 O-O-O) 12 . . . lLlb4 ( 12 . . . �b4t 13 .c3 lWxh4 14 . lLlxh4 0-0-0 Joecks - Arnold, Germany 1 989. Now strongest was 1 5 .if5!;!; and Black has some compensation, but not enough.) 1 3 .0-0 lLlxd3 was played in Paavilainen - Westerinen, Finland 199 1 among others. Now White has an improvement in 14 .lW a4 t! which forces a nice endgame 14 . . . �c6 1 5 .�xc6t bxc6 16.cxd3;!; where only White has winning chances. The key

idea is to play lLl f3-d2-e4 or c4 when the talk of the two bishops will stop.

l 1 . . ..i.f5 12 .ic4 0-0 1 3 .0-0 ixc2 ( 1 3 . . . .i.xd6 14 .ib3±) 14 .if4 �xb2 1 5 . l:1ac1 .i.a3 1 6. l:1fe l ± Doggers - Mek, Tilburg 2003 .

1 1 . . .0-0 12 . .i.d3 lWb4t 1 3 .c3 �xh4 14. lLlxh4 :gad8 was played in Jurek -Suchon, Poland 1 999. Now the strongest continuation is 1 5 .if4 ixd6 1 6 . .i.xd6 :gxd6 1 7.0-0-0± and there is no compensation for the pawn.

1 1 . . . lLlb4 1 2.lWe7 mate! 1-0. Paaske Pedersen, Copenhagen 2003. 12 . .i.d3

1 2.d7t? @xd7 13 .ie2 as in Ribeiro -Fernandes, Lisbon 1 998 only helps Black as 1 3 . . . g5! is very strong now. Other alternatives are equally senseless. 12 .. . 0-0-0

1 2 . . . g5 1 3 . .i.xg5 !? fXg5 14 .lLlxg5 �a5t 1 5 .c3 1-0. Schwab - Dumitriu, corr. 1 995 . 13.0-0


13 . . J�xd6 1 3 . . . g5 14.�h6 g4 1 5 . lLl g5 i.d5

16. lLl e4+- Seirawan - Seybold, Zurich (sim) 1 9S5 .

1 3 . . . h5 14 .i.e3!? Not the only way to play, but a very reliable move. 14 . . . i.xe3 1 5 .fxe3 'lWxe3t 1 6. mhl i.g4 This is given as compensation in ECO, which is completely wrong. Though the book is a very well structured reference guide, it is sometimes too apparent that the lines are made up of a reshuffling of all the games in the Informants, and no evaluation by the editors themselves. This is also why a weaker player's recommendations are made to seem more important than Kasparov's choices over the board, when given the main lines. (Also after 1 6 .. Jhd6 1 7.�g3! Black is in deep trouble. White simply wins a pawn without any counterplay. 17 . . Jl:hdS I S .'lWxg7 i.g4 1 9J1:ae l 'lWb6 20. lLld2 !± .) 17 .�g3!± . A very powerful move. Now Black cannot free himself. 17 . . . 'lWc5 I S . gadl mbS 19 .i.e4 lLl e5 20. lLlxe5 i.xdl This was Hansson -Fernandes, London 1 9S4. Now 2 1 . 'lWxg7 just wins outright. 2 1 . . .'lWcS (2 1 . . .'lWb4 22. lLl c6t! +-; 2 1 . . . �b 5 22.c4+-; 2 1 . . . 'lWb6 22.'lWxhS+-) 22.i.f5! and Black has no squares for the queen since 22 . . . 'IW c5 23. lLl d7t is all over. 14.a3!

Black's pieces are not ready to meet the advance of the queenside pawns. 14 ... lLld4

14 . . . g5 1 5 .�h6! A standard move

in these lines. 1 5 . . . 'lWdS 1 6.b4 i.b6 17 .i.e3± Lamprecht - Bach, Hamburg 1 995.

14 . . . a5 does not work at all. 1 5 .b4! axb4 1 6.axb4 and Black loses material: 1 6 . . . i.xb4 17 .i.e3 i.c5 I S . gaSt lLl bS 19 .i.f4 g5 20.i.xd6 gxh4 2 1 . gxbSt md7 22.i.xc5 'lWxc5 23. gxhS+-. 15.�g3!?

This subtle move is probably even stronger than 1 5 . lLl xd4 i.xd4 16 .'lWg3t when I am not certain that Black does not have a little compensation. Now weak is 16 . . . gd7?! 1 7.i.e3 i.xe3 I S .'lWxe3 'lWxe3 19 .fxe3± as in Negri - Elburg, e-mail 2000. Why did Black want to play this endgame?

15 ... �c6 The alternatives are not better.

1 5 . . . lLl b3? ! 1 6.cxb3 gxd3 1 7.b4 i.d6 I S .i.f4± .

1 5 . . . g5 1 6.i.e3± . 1 5 . . . lLlxf3t 16 .'lWxf3 i.d4 17.c3 i.e5


18 .if5 ± This all seems very clear-cut to me. 16.tLlxd4 gxd4

1 6 . . . ixd4 17 .ie4! id5 1 8 .if4 (also possible is 1 8 .if5t!? ie6 1 9JMxg7 ge8 20. �xh7 and why should Black have compensation for this sea of pawns? 20 . . . gd7 2 1 . �h3 ixf5 22.�xf5 gg8 23.g3+-) 1 8 . . . gd7 1 9.!xd5 gxd5 20.c3+-. This is the simplest. There is no hope for Black here. 17.h3 g5 18.M .ib6 19 . .ib2 gf4 20.gae1

20 . . . .td7?! Black is looking for excuses. I have not

found sufficient compensation for him here, and I cannot see why I should.

The most obvious line goes 20 . . . gd8 2 1 . ge2! h5 22. gfe l h4 23.�xf4 Forced. (23 . �h2? leads to 23 . . . !c4 24.icl ixd3 25.cxd3 gfd4 26.ie3 ic7 27. �hl gxd3+) 23 . . . gxf4 24. gxe6 gd6 25. ge7± and Black's defence is just a nightmare. If

White is actually already winning here is hard to tell, but he has all the chances at least. 21 ..ic1 gh4 22.e4± .te7 23.f4 gd8 24.ga Wb8 25.e5 �d5 26 . .tf1 Ylrd4t 27 . .ie3 �b2 28.gd1 �e2 29.�e1 h6 30.b5 1-0

264 Minor Lines

Minor lines

- By John Shaw

In this chapter we will have a brief look at some of Black's less popular lines. This does not necessarily mean that they are bad, though some are, just that they are played less often. We will in turn look at 2 . . . b6, 2 . . . a6 (O'Kelly) , the Andersson line, the 2 . . . g6 3 .d4 .tg7?! line, different versions of . . . cxd4 followed by . . JMib6, the dubious Lowenthal and finally 1 .e4 c5 2 .lLlf3 e6 3 .d4 cxd4 4.lLlxd4 lLlc6 5 .lLlc3 d6, which is by no means bad at all.

The first game is with 2 . . . b6, which is more respectable than it looks.

Game 61 Thorhallsson - Balinov Plovdiv 2003

1 .e4 c5 2.lLla b6 Afewyears ago this was almost unknown

but now it is a recognised, if still minor, variation. Several grandmasters have tried it and it is even the main defence to 1 .e4 of the Austrian GM Stanec. Here we see it employed by his grandmaster compatriot. 3.d4 cxd4 4.lLlxd4 ib7 5.�c3 a6!?

This move is the reason for the minirevival of 2 . . . b6. Black's idea is to delay the development of his kingside until he has one more move's information about

the destination of White's king bishop. Specifically, if White plays 6.id3 Black will choose a set-up with g7 -g6 and d7-d6. Who would play id3 against the Dragon? While if White tries 6 .ic4 then Black will play e7-e6 and leave the dpawn alone. There are no good ic4 lines against the Kan. White's task is to find a set-up that is threatening against both Dragon and Kan style positions.

The Brazilian 1M Limp has a weakness for the move 5 . . . ltJf6. The simple 6.e5 must be reasonable but 6.ig5 is great fun. His countryman GM Milos played this in his second attempt against 5 . . . ltJ f6, so we can assume he prepared 6.ig5 and trusts it. After 6 . . . ltJxe4 7. ltJxe4 ixe4 S .'!Mfe2 White has a dangerous lead in development. Milos -Limp, Sao Paulo 2004, continued S . . . d5 (S . . . ib7? illustrates the danger. 9 . ltJb5! and White is already completely winning. For example, 9 . . . d5 1 0.if4, or 9 . . . d6 1 0 .0-0-0.) 9 . E!d1 h6 10 .ih4 a6 l 1 .f3 ih7 12 . 'it> £2 (Intriguing, but I would prefer the immediate 1 2.c4! . ) 1 2 . . . Wfd7 ( 12 . . . b5!?) 1 3 .c4 and White had a strong initiative and soon won. 6.ie3!

I think this is the logical answer and guarantees White an edge. White plans '!Mfd2, £2-f3 and 0-0-0. Whether this is a Yugoslav Attack or an English Attack is up to Black.

I also like the look of 6 .ig5!? This is not quite as clear as 6.ie3, but it does

give White the chance to play for mate at a shockingly early stage of the game.

Now after 6 . . . h6?! the weakening of g6 is more than just theoretical. 7 . .ih4 d6 8 . .ic4 � c8 (This may seem odd but 8 . . . � c7 allows the incredibly annoying 9 . .ie6! . The justification is 9 . . . fxe6 1O .ltJxe6 �c8 1 1 .�h5t @d7 12.�f5 and White wins.) 9 .�e2 ltJd7 10.0-0-0 g6 1 l .f4 .ig7 1 2.e5 dxe5 was Moreno Camero - Komeev, Dos Hermanas 2003. Now White missed his big chance with 13 . ltJe6!! when Black has no defence. For example, 1 3 . . . fxe6 14 . .ixe6 .ic6 1 5 .�g4.

The punch of 6.!g5 is shown by the fact that a 2600 player was lost with Black after only 1 3 moves.

6 . . . � c7 prepares e7 -e6 and is probably the only move to avoid immediate trouble. Now White has several reasonable moves such as 7 . .id3 or even 7.\Wg4 but a critical try is 7. ltJd5!? when Black has little choice. 7 . . . ,ixd5 (7 . . . �e5 8 . .ie3 should just be a transposition after 8 . . . ,ixd5, but in Groszpeter - Berezjuk, Pardubice 2000 Black blundered with 8 . . . e6?? when 9 .1tJf3! won easily.) 8 .exd5 �e5t 9.!e3 �xd5 1O . .ie2 I believe White's massive lead in development gives him excellent compensation for the pawn. 6 ... 00

This is now the only sensible set-up. Playing in Dragon-style would be mad with the passive a6 and b6 already played, and White heading for a Yugoslav Attack.

7.V!Yd2 J.b4 8.S <tf6 Instead 8 . . . \Wc7 transposes to Karjakin

- Balinov, Vienna 2003 . After 9.a3 .ie7 (If 9 . . . .ixc3 then either recapture is good enough for an edge. Black will have weak dark squares and no bishop to cover them.) 1 0 .0-0-0 b5 1 1 .!f4 \Wc8 12 .ltJb3! ltJc6 13 . .id6 Black's position was disgusting. This game was played two months before our main game, so we can assume that the text is GM Balinov's (unsuccessful) attempt to improve. 9.a3 ixc3

Now Black has problems on the dark squares but 9 . . . .ie7 1 0.e5! was even worse. For example, 1 0 . . . ltJd5 l 1 . ltJxd5 .ixd5 12 .1tJf5 ! is a disaster. 10.V!Yxc3 d5 l l .e5 <tfd7 12.0-0-0 <tc5 13.f4

White has an excellent version of the Classical French. 13 ... V!Yd7 14.<tS

Challenging Black's only well-placed

266 Minor Lines

piece with 14 . tD b3! seems more logical. For example, 14 . . . tD e4 1 5 .�b4 b5 16 .!d3 tD c6 17 .�e1 and White has a great position. 14 ... 0-0 lS.J.d3 gc8 16.�h1 as

Planning to improve his feeble bishop with ia6, but the tactics dictate that tDxd3 must be played very soon. 17J�hel Aa6?

This was Black's last chance to reach a decent position with 17 . . . tDxd3. 18.hcs

White begins to clear the queens path to h3. 18 ••• hxcS

It was too late for 1 8 . . . ixd3 because of 1 9.�xd3 bxc5 20. tDg5 g6 2 1 . �h3 h5 22. tD e4 with a winning attack.

19.J.xh7t! The Greek Gift is still claiming

victims. 19 .•• �xh7 20.tDg5t �g8 21 .Y!Yh3

White is winning since f4-fS will

further sttengthen the attack. 21 ••• Y!Yd8 22.Y!YhS ga7

Or 22 . . . �e8 23.fS! tDd7 24.�h7t WfB 25. tD e4! and Black has no defence. For example, 25 . . . dxe4 26Jhd7 �xd7 27.f6 leads to mate. 23.5! ex5 24.e6 g6 2S.Y!Yh7t �f8 26.e7t 1-0

The O'Kelly Variation has never been regarded as fully respectable, yet several GMs are willing to risk it. Former World Junior Champion Kurajica is a true believer and the 2600-rated BakIan has recently added it to his repertoire. Still, White should get an edge.

Game Shirov - Kurajica Sarajevo 2002

1 .e4 cS 2.�f3 a6 3.d


I hesitated before selecting this as our main line: a c3-Sicilian in a book that recommends only the most critical lines? I convinced myselfby studying the database: 3.c3 is the anti-O'Kelly choice of even the most aggressive of the world's top GMs (Shirov, Bologan and even Kasparov) . It is also the move that scores most heavily against the 2 . . . a6 experts. The convincing logic of 3.c3 is that a7 -a6 is rarely a useful move in the normal c3-Sicilian.

Those determined to play in Open Sicilian style could consider 3 .ltJc3. After 3 . . . b5 4.d4 transposing to our anti-Kan repertoire may seem likely but Black has enjoyed considerable success with the surprising 4 . . . e6. The critical line should be 5 .d5 but then we reach a 1 .d4-style position where White seems to have "forgotten" to answer . . . a6 with a4.

3.c4 is a logical, reliable move, but if Black chooses a line with e7-e6 the positions reached are inconsistent with both this book's anti-Kan/Taimanov recommendations and our attacking piece play style. Furthermore, the 2 . . . a6 specialists have had several recent successes in the line 3 . . . d6!? 4.d4 ig4. The idea is if 5.dxc5 then 5 . . . ixf3 and Black hopes to land a mighty knight on d4.

The trick every player must know (and probably already does) is that 3.d4?! is a mistake. After 3 . . . cxd4 4. ltJxd4 ltJf6 5 .ltJc3 e 5 Black has an excellent version of the Sveshnikov with ic5 or ib4 as extra options. The usual ltJdb5 is ill

advised and 6. ltJf5 d5 is not clever either. Instead White would have to retreat the knight passively, and hope to hang on to equality. 3

. . . g6 This is one of many possible replies. The

unforced nature of the play means that it is neither possible nor necessary for White to memorise much theory in this position. Black has a multitude of ways to achieve a worse version of a normal variation. For example:

a) The most solid try is probably 3 . . . d5 but after 4.exd5 �xd5 5 .d4 White's "extra" move gives him a pleasant position. One example: 5 . . . ltJf6 6.ie2 cxd4 7.cxd4 g6 8 .ltJc3 Y;l(d8 9.0-0 ig7 I O .ltJe5 0-0 I 1 .iS ltJe8 12 .if4 ltJd6 I3 .Y;l(d2 ltJf5 I4J'!:adI ltJd7 I 5J'!:feI and White was clearly better in Karjakin - Khamrakulov, Dos Hermanas 2004.

b) 3 . . . e6 4.d4 d5 5 .e5 gives White a good version of the Advance French. Black can try to make use of 2 . . . a6 with 5 . . . id7 planning to exchange the bad bishop on b5 but this runs into problems. For example, 6.id3 (Not the only good move. Reasonable alternatives include 6.a3 and 6. ltJ bd2.) 6 . . . cxd4 (The problem with 6 . . . ib5 is 7.ixb5t axb5 8.dxc5 ixc5 9 .b4 ib6 1O .ltJa3 and White wins a safe pawn.) 7. ltJxd4! . Cutting across Black's plans. 7 .. . lt k6 8. ltJxc6 ixc6 9.0-0 and White's lead in development gives him attacking chances on the kingside.

268 Minor Lines

4.d4 cxd4 5.cxd4 d5 6.exd5 �f6 7.�c3 �xd5 8 . .ic4 �xc3 9.bxc3 ig7 10.h4!?

This is certainly not forced, but it does exploit the slowness of Black's second move. If White tried to hack a normal c3-Sicilian in this style then Black would never consider the irrelevant a6 in reply. 10 . . . Wc7 1 1 .Wb3 e6 12.id3 �c6 B.h5 id7 14.�g5 gc8 15.�e4 0-0 16.hxg6 hxg6 17.ig5 f5

Black should probably have fought for control of the dark squares with 17 . . . f6 , but White still has good attacking chances. 18.�c5 �a5 19.Wb4 gf7 20.gbl i£8 21 .'it.>f1 ic6 22.'it.>gl e5 23.dxe5 Wxe5 24.ie3 Wd5 25.if1 Wxa2 26.Wh4 gg7 27.gd1

Among many, many others, Black has to defend against 28. lt:lxb7. 27 .. .id5 28.gd2

Winning the queen for rook and minor piece is good enough, but White could

have concluded the attack with 28. 'lWh8t mf7 29. gh7! gxh7 30.'�xh7t !g7 3 1 .!h6 gg8 32.c4! . 28 .. . Wxd2 29 . .bd2 gxc5 30.Wd8 gf7 31 .!e3 �c6 32.Wh4 gg7 33 . .bc5 hc5 34.ic4 1-0

The following sideline, here ascribed to Ulf Andersson, is sometimes called the Haberditz variation.

1 .e4 c5 2.�f3 �c6 3.d4 cxd4 4.�xd4 �f6 5.�c3 e5 6.�db5 h6

This rare move is a favourite of GM Ulf Andersson. Most of the main lines involve an early exchange of queens, which explains the interest of this famous endgame expert. However, White is able to achieve an edge with accurate play. 7.�d6t

This is the logical way to exploit the omission of d7 -d6, but there are other


options. 7. itJ d5 should be avoided. After 7 . . . itJxd5 8 .exd5 a6! analysis and practice show that Black has equalised. For example, 9 .dxc6 axb5 1 O.cxd7t i.xd7 I I . �d5 �c7.

If White is determined to keep the queens on then 7.i.e3!? is worth considering. Black must play 7 . . . d6 8 .itJd5 itJxd5 9 .exd5 itJ b8 . The only move. (9 . . . itJ e7? loses a pawn to 10 .c3! because if 10 . . . a6?? then 1 1 . WI a4!, a standard trick known from the 7. itJd5 Sveshnikov.) 1 0.c4. Compared to a normal 7. itJ d5 Sveshnikov White has the extra move i.e3, and Black has . . . h6. Generally, this must favour White, but . . . h6 may help prepare a later i.e7-g5 . I would recommend this line to players who prefer an unclear middlegame to a slightly better ending. 7 . • .J.xd6 8:�xd6 We7 9.Wxe7t

This is the simple approach. The sharpest try is 9 . itJ b5! ? but Black has recently strengthened his defence with 9 . . .l':!:b8! . This obviously avoids the fork but also fights to prevent the knight landing on d6. Previously 9 . . . Wlxd6 was always played with White scoring well after 10 . itJxd6t me7 I l . itJ f5t. After 9 . . . l':1b8! one critical line is 1 O .b3 itJxe4! 1 1 . �xe7t mxe7 12 .i.a3t d6 1 3 .8 a6! 14.fxe4 axb5 1 5 .i.xb5 ga8 and Black is fine. 9 ... cbxe7 10.J.e3 d6

10 . . . itJ g4? is premature. I 1 . itJ d5t md8 12.i.c5 and Black is crushed.

1 1 .£3 J.e6 12.0-0-0

White has a slight but definite edge. The plan is gl-g4 and h2-h4 preparing a later g5 . White can wait and play itJ d5t only if and when it is most inconvenient for Black.

The following line has been around for a while, but it does not seem to be fully trustworthy.

1 .e4 c5 2.tLl£3 g6 This is a perfectly acceptable move

order to reach either variety of Dragon. However, if Black avoids this possibility then he runs into immediate trouble. Of course White has third move alternatives, but none of them are superior to 3.d4. 3.d4 .tg7?!

Black can, and should, head back towards regular Dragon lines with 3 . . . cxd4 4. itJxd4. White can prevent a transposition with 4. Wlxd4 but I think

270 Minor Lines

Black is OK after 4 . . . lLlf6. For example, 5 .i.b5 a6 6.e5 axb5 7.exf6 lLlc6. 4.dxc5

It is worth noting that 4.c4 does not force a transposition to the Accelerated Dragon. Black has interesting alternatives in 4 . . . �b6 and 4 . . . �a5t. 4 ••• Wa5t 5.d!

5 . lLl c3 is a little speculative. After 5 . . . ixc3t 6.bxc3 �xc3t 7.i.d2 �xc5 Black has had encouraging results. 5 •.• Wxc5

6.tLla3! This causes Black far more problems

than the obvious 6.i.e3. Black now has six likely replies of which four are clearly bad. 6 ••• tLlf6

a) 6 . . . i.xc3t? is a trap which has caught several strong players. 7.bxc3 �xc3t 8.�d2! �xal 9 .lLlb5 . The threatened fork gives White just enough time to trap the queen. 9 . . . lLla6 1O .lLlc3. There is no

escape. For example 10 . . . lLlc5 l 1 . lLl d4! and White wins easily.

b) Preventing lLlb5 with 6 . . . a6? is disastrous after 7. lLlc4 with i.e3 to follow.

c) 6 . . . � a5?! 7. � d5! forces an advantage in every variation: 7 . . . lLl c6 8. �xa5 lLlxa5 9 .lLlb5, 7 . . . �d8 8.i.c4 e6 9 .�d3 and 7 . . . �xd5 8.exd5 lLlf6 9.d6 - Bruwn.

d) 6 . . . � c7?! allows White to gain a clear advantage: 7. lLlb5 �d8 (7 . . . �a5 8 .�d5!) 8 .i.f4 d6 9.e5 a6 10 .�a4 - Bruzon.

e) 6 . . . d6 is the only serious alternative to 6 . . . lLlf6, but White can still achieve a slight advantage. 7. lLl b5 a6 (7 . . . �b6?! wastes more time than Black can afford. Markowski - Danner, Istanbul (01) 2000, continued 8.i.e3 �a5 9 .i.c4 lLlc6 10. �b3 lLld8 l 1 . lLlxa7! . And now 1 1 . . . �xa7? loses to 12 .i.b6.) 8 .i.e3 �c6 9 .lLla7 �c7 1O .lLlxc8 �xc8 1 1 . �b3 lLlf6 12.e5 dxe5 1 3 .lLlxe5 0-0 14.i.e2 and White was a little better and later won in Dolmatov - Solak, Istanbul 2003. 7.tLlb5 0-0

7 . . . lLlxe4? is rather naive. 8 .Be3 �c6 9 .lLlfd4! i.xd4 10 .�xd4 lLlf6 l 1 .lLlxa7 and White has a clear advantage.

7 . . . lLlg4 is direct but White has a convincing answer. 8 . lLlfd4 d6 9 .h3 lLlxfL.. This is effectively forced. (For example, 9 . . . lLl f6? 10 . i.e3 wins.) 10 . cj(xf2 e5 1 1 .i.e3 exd4 1 2.cxd4 �b6. This is AI Sayed - Velikhanli, Abu Dhabi 2003. Now 13 . �cl lLla6 14.e5! is very strong. S.i.e3 Woo 9.tLlfd4 Wxe4 lO.tLlc7 b6


So far this is Bruzon - Malakhov, Yerevan 2000. Now Bruwn's suggestion of 1 1 .Wa! forces a queen exchange while still winning the exchange.

The line with 2 . . . llJc6 and 4 . . . VNb6 is a favourite of Israeli GM Golod, who plays it with absolutely fantastic results (940/0, a fabulous 2790 performance in the last 5 years. But all games were played against lower rated opponents . ) . It should not be underrated. Still, it is not the refutation of the Open Sicilian and White can fight for an advantage just as in any other system. Especially if he follows my recommendation.

Game 63 Movsesian - Kozul Sarajevo 2002

1 .e4 c5 2.ttJa ttJc6 3.d4 cxd4 4.ttJxd4 Wb6

This is an interesting, playable line. Black nudges the d4-knight towards b3, while keeping his own pawn structure as flexible as possible. 5.ttJb3 ttJf6 6.ttJc3 e6

7.We2!? I think this is the best way to attack

Black's system. White prepares queenside castling, while the fl bishop will develop to g2. A common alternative is 7.i.e3 but it not clear that this is a gain of tempo. Firstly, the queen is almost certain to move anyway to enable a later b7-b5. Secondly, since ib4 is a definite possibility, d2 may prove to be the correct square for the bishop. 7 ••. d6

The major alternative is 7 . . . i.b4 S.i.d2 0-0 9.a3 and now Black has a major decision to make:

a) Most strong players have given up 9 . . . i.xc3 1O .ixc3 e5 1 1 .0-0-0 EidS. (In A. Ivanov- Yermolinsky, USA (ch)

272 Minor Lines

2000, Black tried the tricky 1 1 . . . d6. After 1 2J hd6 lLld4 13 .i.xd4 1Mfxd6 14.i.cS 1Mfc7 I S .i.xfB �xfB 16.f3?! (Instead I think 16 .1Mf c4! refutes the gambit. The idea is 16 . . . 1Mfxc4 17.i.xc4 lLlxe4 18 Ji:e l lLld6 (Not 1 8 . . . lLlxf2? 19 J'm) 1 9 .'!d3 and White will soon be a pawn ahead.) 1 6 . . . .!e6 Black had enough for the pawn.) 12 J '!d6! 1Mfc7 ( 12 . . . lLld4?? is now simply a blunder. 13 . !'lxd4 exd4 14 . .!aS .) 13 . !'lxf6! This idea of Michael Adams has effectively ended the popularity of 9 . . . .!xc3. 1 3 . . . gxf6 14 .1Mfg4t �h8 I S .1Mfh4 1Mfd6 (A later game Nijboer -Piket, Amsterdam 200 1 , continued I S . . . d6 1 6.1Mfxf6t �g8 17.1MfgSt �h8 1 8 .f4 and White was much better.) 16.f4 1Mfe7 (Nijboer suggested 16 . . . !'lg8 as a better defence but also supplied a promising reply. 1 7.g4!? 1Mfe7 1 8 . .!bS d6 1 9 .h3 and White has superb compensation. For example, 1 9 . . . �g7 20. !'lfl a6 2 1 .'!xc6 bxc6 22.fxeS dxeS 23 . .!b4 1Mfd8 24. lLlaS ! and Black i s lost.) 17 .i.bS Adams -Knezevic, France 1997. Black is already in trouble since the natural 17 . . . d6? fails to 1 8 . .!xc6 bxc6 1 9.fxeS dxeS 20.i.b4!.

b) 9 ... .!e7 l O.eS. It is wise to push now since 10.0-0-0 dS! has scored exceptionally well for Black. l O . . . lLldS 1 1 .0-0-0 (If I I . lLlxdS exdS 12 .0-0-0 d6 Black's potential c-file play compensates for his weaker structure.) 1 1 . . .d6 12.exd6 i.xd6 13 .lLl bS! .!eS ( l 3 . . . .!f4 was tried in Kotenko - Arzumanian, Ukraine 2003. After 14.c4 .!xd2t I S . !'lxd2 lLlc7 16.cS

1MfxbS 17.1MfxbS lLlxbS 1 8 .i.xbS White had a definite edge.) 14.c4 This only looks risky- "Trust me, I know what I'm doing." 14 . . . a6 I S .cxdS exdS 16 . .!e3 1MfxbS 17.1MfxbS axbS 18 . .!xbS .!e6 Motylev-Kunte, India 2002. Now Judit Polgar suggests 19 . !'lhe l is slightly better for White.

Black can also play 7 . . . 1Mfc7, but this generally transposes to the 7 . . . d6 variations. One independent example is 8.g4 h6 9.h4 a6 l O . .!g2 i.d6. A creative and unusual attempt to stop White's traditional pawn storm. 1 1 .i.e3 ( l l .gS!?) l l . . .bS 1 2.0-0-0 .!f4?! 1 3 .lLldS ! and White was much better in Ciuksyte - V. Georgiev, Porto San Giorgio 2003. 8.g4 a6 9.1d2 fie7 10.0-0-0

Black has far too many possibilities here to give comprehensive coverage. However, White appears generally to have the better chances in an unbalanced position. 10 .•• �d7

Black can also choose to delay g4-gS with l O . . . h6. Now 1 1 .h4 g6 12 . .!g2 bS transposes to Tomescu - Bruno, Porto San Giorgio 2002. Play continued 1 3 .eS! dxeS 14.1Mff3 lLldS I S .lLlxdS exdS 16.1MfxdS .!b7 ( l6 . . . i.d7 17 . .!aS!) 17. lLl cS . Black is already worse but 17 . . . !'ld8? 1 8 .lLlxa6! 1Mfb6 1 9 .1MfxbS! did not help. 1 1 .£4 b5 12.g5 �e5

Kozul clearly believes 12 . . . b4 1 3 .lLla4 safely blocks his queenside attack. 13.�xe5 dxe5 14.J.e3 �d4 15.fifl 1b7 16.J.g2 0-0-0 17.e5


17 . . . .ie7 18 . .ixb7t c;f?xb7?! Now White wins a safe pawn. Black

had to try 1 8 . . . �xb7 but after 1 9J1:he l White i s still a little better. 19 . .ixd4 cxd4 20.gxd4 gxd4

Of course if 20 . . . �c5 2 1 . Wff3t. 21 .�xd4 .ie5 22.Yfd3 gd8 23.Yfat Yfc6 24.gfl gd4

There is still work to do but Movsesian completes the job convincingly. 25.�e2 ge4 26.�g3 ga4 27.c;f?bl Yfxa 28.gxa .igl 29.c3! ge4 30.c;f?c2 ge7 31.�e2 .ib6 32.gh3 c;f?e6 33.gxh7 g6 34.c;f?d3 c;f?d5 35.h4 a5 36.�g3 b4 37.�e4 bxc3 38.bxc3 ge4 39.�f6t 1-0

2 . . . e6 and 4 . . . �b6 is a rare line now but an old favourite of GM Kveinys (which explains why a later . . . Wfb6 in the Kan is known as the Enhanced Kveinys variation) . The idea, as in many of the Kan lines, is to encourage the .!L\ d4 to leave its perfect central position.

Game 64 Yu Shaoteng - Zaw Wm Lay Bangkok 2004

1 .e4 e5 2.�a e6 3.d4 cxd4 4.�xd4 Yfb6

4 . . . ic5 is likely to transpose back to our main game after 5 . .!L\ c3 Wfb6, but White can also try an immediate attack on Black's position with 5 . .!L\ b5! ? Wfb6 6.ie3.

First played by Paul Morphy! It gives White good compensation and a wonderful score after 6 . . . ixe3 7.fxe3 �xe3t. The only move. (7 . . . �e7 was played once, but 8. � d3! with the ideas �b4t and Wfc3, as well as 9.e5 blows Black away. One line is 8 . . . d6 9.e5! dxe5 10 .Wfa3t �f6 1 1 . .!L\ d2 .!L\ c6 12.id3 with the makings of a winning attack. e.g. 12 ... �xe3t 13 . �dl .!L\ h6 14Jl:f1 t .!L\ f5 1 5 . .!L\c7 l:l:b8 16 . l:l:xf5t! and White wins. And after 7 . . . .!L\ f6 8 . .!L\d6t �e7 9 . .!L\ c4

274 Minor Lines

WEc5 10 .e5 tLl e8 ( l 0 . . . tLld5 1 1 .'iNh5! and I do not see how Black can defend his kingside.) I l . tLl c3± Hruciov - Rusev, Oropesa del Mar 2000.) 8 .�e2 tLl a6 9. tLl l c3 tLlh6 (9 . . . tLlf6 10 .tLld6t mf8 1 1 . EW White has far more compensation than needed for the pawn, Staniszewski - Ostrowski, Augustow 1 996.) 1 0. tLl d6t me7 I l . tLl c4 'iNc5 1 2.a3 b5 1 3 .b4 WEc6 14 . 'iN d4 gave White a wonderful initiative in Brodsky - Staniszewski, Rowy 1 999. He converted this to a better endgame, but a full-blown attack on the king was

also an option. 5.tLlc3

The most common move is 5 . tLl b3 but in an active repertoire it is worth avoiding this slightly passive retreat if there is a reasonable alternative.

The unusual 5 .tLla3!? was once played by the creative Russian GM Dvoirys. In the game Dvoirys - Agrest, Cattolica 1 994, Black allowed the planned tLl c4 and was soon worse. The critical try must be 5 . . . .!xa3 6.bxa3 when the positional question is obvious - will White's dark square play be more important than his weak pawns? Dvoirys gave the line 6 . . . tLl f6 7.'iNd3 0-0 8 .�e2 tLlc6 9 . tLl b5 d5 1 O .exd5 exd5 1 1 ..!b2! tLle4 1 2.0-0 .!e6 13 . Eiab l a6 14. tLld4 tLlxd4 1 5 . .!xd4 WEc7 1 6.a4± which is interesting but not forced. 5 ••• Ad 6.tLla4

If 6.ie3 Black is OK after 6 . . . tLlc6, but he must avoid 6 . . . 'Wxb2?? 7 .tLldb5;

winning instantly. 6 ••• 'ilYa5t 7.c3

7. tLlc3 WEb6 is relevant only as a way of playing out a pre-arranged draw. 7 •• .J.xd4

Black is committed to giving up his better bishop as 7 . . . �e7 8 .tLlb5 d6 9 .�f4 e5 1 0 . .!e3 is simply good for White. 8.'ilYxd4

Basically, White already has what he wants - the bishop pair. The onus is on Black to find compensation. 8 ••• tLlf6

Or 8 . . . e5 9 . 'iN c5 and White keeps a clear edge. 9.tLlc5 tLlc6

9 . . . 0-0 1 0. tLl b3 allows White to consolidate his gains. lO.'ilYd6!?

I prefer this positional pawn sacrifice to the more common and passive 1 0.WEe3, which seems to give Black good equalising chances.


10 ... b6 1 1 .�b3 YlYe5 It is telling that Zaw Win Lay switched

to this move after trying the main line a year earlier. The critical position is reached after 1 l . . . ttlxe4 12 ."%!fxc6! dxc6 1 3 .ttlxa5 bxa5 14 .1e3.

Black has a healthy extra pawn on the kingside, but his shattered queenside and White's bishop pair constitute more than adequate compensation. I believe that White's control of the position means that he can play for the win with virtually no risk of defeat, or in modern jargon: White is playing for two results.

Hamdani - Zaw Win Lay, Vietnam 2003, continued 14 . . . 0-0 (Instead, Nikolenko - Arzumanian, Tula 2000, continued 14 . . . e5 1 5 .0-0-0 1e6 1 6.1a6 Eibs 17 . Eihe 1 ttl f6, and now the simple 1s .ic5 would have given White a clear advantage.) 1 5 .0-0-0 e5 1 6.1c4 1f5 17.8 ttl f6 1 S . Eid6 EifcS 19 . Eihd 1 . This is a typical position in this variation. White

dominates the d-file while Black is tied to defending his weak pawns. 1 9 . . . ttl eS 20. Ei6d2 ttl c7 and now White, the lower rated player, headed for a draw with 2 1 . :Bd6. Instead he could have played for the win with no risk in several ways. One example is 2 1 .a4!? , simply fixing the weaknesses. 2 1 . . .1e6 22.ixe6 ttlxe6 23. Eid7 and White will soon recover his pawn with a fine position. 12.YlYxe5 �xe5 13.£4 �g6 14.e5 �d5 15.g3 ib7 16.J.g2 0-0-0 17.0-0

Without doing anything special White has achieved a slight but definite edge. 17 ... d6 18.exd6 gxd6 19.�d4 �d7 20.a4

It was worth considering 20.f5!? exf5 2 1 . ttlxfS f6 and only then 22.a4. 20 ... �de7 21 .a5 hg2 22.�xg2 �b7 23.ie3 �c6 24.axb6 axb6 25.�xc6 �xc6 26.�a6 gb8 27.c4 �e7 28.b4 �f5 29.igl gd3 30.c5 �e3t 31 .ixe3 gxe3 32J�c1

The simple 32. Eid1 gives White decent winning chances. 32 .. J!b7 33.cxb6t �b5 34.�cal gc3 35.�la5t �xb4 36.ga7 gc2t 37.�h3 �xb6 38.�al f5?!

3S ... e5 should draw. 39Jhg7 �c5 40.gc7t gc6 4U�xh7 ge2

Black has a very unpleasant defensive task ahead, and soon cracks. 42.ghl �d5 43.�h4 �e4 44.�g5 �f3 45J'Ial ge3 46.ga5 �g2 47.ge5 ge4 48.gh6 �a6? 49.�xf5 1-0

276 Minor Lines

The next line we will examine is the Lowenthal, which was first played more than 100 years ago. It has never achieved any popularity in high-level chess. Black's weakened dark squares do not appeal to strong players. At lower levels the Lowenthal is far more common. I suspect the large number of tactical tricks in the main lines is the main attraction.

1 .e4 c5 2,c�:lf3 �c6 3.d4 cxd4 4.�xd4 e5 5.�b5 a6

This is the initial position of the LOwenthal. 5 . . . d6 is of course the Kalashnikov (see page 2 1 9) . 6.�d6t hd6 7.Wxd6 Wf6

8.Wxf6 This is an easy way to guarantee an edge.

The positions tend to be not very exciting, but they are even less fun for Black. The main line is considered to be S .Wd1 but I do not think it is any better than S. �xf6. It also commits White to studying and

remembering a large amount of theory about a rarely met line. 8 ... �xf6 9.�c3

Black is now committed to searching for compensation for his positional concessions. 9 ... ll:Ib4

The only other significant try is 9 . . . d5 . Now 10 .exd5 is better than the frequently recommended 1 0.i.g5 . After 1 0.exd5 play continues 1 0 . . . � b4 1 1 .i.d3 �xd3t (If 1 l . . .ttJfxd5 12 . ttJxd5 ttJxd5 13 .i.d2 White has an edge in a simple position.) 1 2.cxd3 i.f5 13 .0-0!? (I find this clearer than the also promising 13 .i.g5 . ) . Now:

a) 1 3 . . . 0-0-0 14 .ig5 hd3 1 5 .E:fd1 if5 16.E:ac1 and Black is in trouble. 1 6 . . . �d7 ( 1 6 . . . �bS 17.f4) 17.�a4.

b) 13 . . . 0-0 14 J�e1 gfeS 1 5 .i.g5 i.xd3 1 6. gad1 i.f5 17 .i.xf6 gxf6 1 S .d6.

c) 1 3 . . . ixd3 and now 14 . ge1 wins a pawn. 10.<bd2 d5 This is the only aggressive try. The quiet 10 . . . d6 changes nothing. A sample line: 1 1 .a3 ttJc6 12 .i.d3 i.e6 13 . gdl 0-0-0 14 . <j;Je1 and, as usual, Black will suffer in a long ending. 1 l .a3 d4 12.axb4 dxc3t 13.<be3

This convincing line has been known for decades. 13 ... �g4t

The critical attempt but White has it covered. Quiet play will leave White with a simple advantage. One recent example is 1 3 . . . i.e6 14 .i.d3 0-0 1 5 .f3 gacS 1 6.b3


lLle8 17.b5 lLlc7 1 8 .bxa6 bxa6 19 .i.a3 Eid8 20.i.b4 and White was already winning in Kotronias - Stankovic, Greece 2002. GM Kotronias is not a player who ducks a theoretical challenge, so he clearly believes 8. 'I1*fxf6 is an effective answer to the Lowenthal. 14.�e2 f5 15.bxc3 c!lJf6 16J:�a5!

and White has a clear advantage.

In the last game of this chapter we shall look at an anti-Keres Scheveningen line. Naturally I suggest playing the Keres attack all the same.

Game 65 Senff - Schlosser Germany 2004

1.e4 c5 2.c!lJa e6 3.d4 cxd4 4.c!lJxd4 c!lJc6 5.c!lJc3 d6

This has never been a popular position with Black players though there is nothing

particularly wrong with it. White has many options, but the consistent move with our anti-Scheveningen line, and also the most theoretically respected, is 6.g4!

The Pseudo-Keres Attack seems like a reasonable name. This move first attracted attention after being played by Anatoly Karpov in his 1 985 World Championship match against Kasparov.

Instead 6. i.e3 is a good alternative but 6 . . . lLlf6 transposes to a Scheveningen line outside this book's repertoire, so we will keep our focus on 6.g4. 6 ... a6

This standard Sicilian move is the most popular, but Black has several reasonable, and little explored, alternatives.

6 . . . h6 was Kasparov's choice in his previously mentioned game against Karpov. Karpov continued 7.h4, but the developing 7. i.e3 is also fine. Transpositions are always possible but one distinct, and inspirational, example is Kasimdzhanov -van der Sterren, Germany 200 1 . 7.i.e3 lLl f6 8 .h3 i.e7 9 .'lWd2 a6 1 0.0-0-0 lLlxd4 1 1 .'lWxd4 i.d7 1 2.f4 i.c6 1 3 .Eigl lLld7 14 .i.c4 'I1*fa5 1 5 .i.xe6 fxe6 1 6.'lWxg7 i.f6 17 .'lWg6t me7 1 8 .e5 dxe5 19 . Eixd7t mxd7 20. Eidl t me7 2 1 .g5 Eihf8 22. 'I1*fh7t Eif7 23.gxf6t mxf6 24.'lWxh6t me7 25.f5 1-0.

6 . . . i.e7 is also reasonable and may transpose to other lines. This really is an ideal variation for those who would rather play chess than learn theory.

278 Minor Lines

Of course 6 . . . ltJ f6 is a regular Keres Attack. See Page 1 99 . 7.J.e3 tLlge7

Now the game is distinct from the real Keres. Black plans to ease the congestion by ltJxd4 followed by ltJ e7-c6. 8.tLlb3!

This is the key point to remember. The idea is borrowed from the Taimanov Variation. ltJd4-b3 is usually a passive move but here it leaves Black's pieces cramped and uncoordinated, in particular Black's knights, which are on the same circuit and so "step on each other's toes." 8 ... b5

9.Wd2 I would prefer 9.f4 first. The possible

downside of this move order is the pawn sacrifice 9 . . . g5, but after 1 0.fxg5 I do not believe in Black's compensation. The knight will look very pretty on e5, but White also has active pieces and Black's king has no safe haven.

9 .a3!? is an interesting and unusual way to avoid the regular lines. Gallagher - Klauser, Switzerland 2003, continued 9 . . . ltJg6 10 .g5 !b7 l 1 .h4 ltJge5 1 2.f4 ltJc4 13 .J.xc4 bxc4 14 . ltJd4 with an unclear position. 9 ... id7?!

The normal move here is 9 . . . !b7. Svidler - Bischoff, Bled (ol) 2002, continued 1 O .f4 ltJc8 1 1 .0-0-0 !e7 12. mb1 0-0 13 .g5 with a sharp opposite side castling position. In other words a fairly typical Sicilian.

The move I want to avoid with the 9.f4 move order is 9 . . . ltJ e5 with unclear play. 10.0-0-0 tLlc8 1 1 .f4 te7 12.g5 0-0 13.e5!

Black suffers because of the position of his bishop on d7. 13 . . .!e8?

This allows White to decisively strengrhen his kingside attack but Black's position was already unpleasant. For example 1 3 . . . ltJa5 14 . ltJxa5 YMxa5 1 5 .YMg2!? :gb8 16 .mb1 and White is clearly better. 14.tLle4 dxe5?!

Now White has a forced win but the alternative 14 . . . d5 still leaves Black in trouble. 1 5 .YMg2 mh8 (or 1 5 . . . YMc7 1 6. ltJ f6t mh8 17.J.d3) 1 6.YMh3 and White's simple, crude attack is deadly. 15.Wg2 We7 16.tLlf6t!

Experts vs. the Sicilian

16 . . . 'it>h8 17.i.d3 J.xf6 If 17 . . . exf4 White wins with 1 8 .tlJxh7!.

For example 1 8 . . . fxe3 19 .tlJf6! and mates.

After 17 . . . gxf6 White has an easy win with 1 8 .gxf6 �g8 (if 1 8 . . . ixf6 then 1 9.�e4 and mate next move) 1 9 .�h3 �g6 20.ixg6 fxg6 2 1 .�xe6. 18.gxf6 g6

Black is forced to weaken his structure since 1 8 . . . �g8 loses immediately to 1 9 .ixh7!. 19.tfg5 exf4 20..bf4 e5 21 .h4!

Not the only way to finish but definitely the most stylish. 21. .. exf4 22.tfh6 gg8 23.h5

Black has no answer to the beautiful threat of 24. �xh7t! @xh7 25.hxg6 mate. 1-0

279

280 5th move alternatives

5th move alternatives

- By Jacob Aagaard

In this chapter we shall investigate Black's alternatives on the 5th move to the normal lines. In the following position

Black has some alternatives to the normal 5 . . . a6, 5 . . . g6, and so on. These include 5 . . . e5, 5 . . . lLld7 and 5 . . . .id7. Though none of them are really completely reliable they are still not as bad as might be imagined. In this chapter I will quickly present a way to play against each of them.

5 . . . e5

This line is a true provocation. Normally Black plays 5 . . . a6 in order to play . . . e5 without allowing .ib5t, but here Black decides to allow it. This will lead

to a position where White has a slight positional pull, and the better player will most likely win with White, and draw with Black.

Basically the 5th move alternatives presented in this chapter have their drawbacks, but are not really bad moves as such. Therefore it is usual for White to achieve a slight advantage, but not more.

1 .e4 c5 2.�f3 d6 3.d4 cxd4 4.�xd4 �f6

Lately 1M Bator has gone 4 . . . e5 5 . .ib5t lLld7 6. lLlf5 a6 7 . .ixd7t �xd7 when he is retaining some flexibility with the g8-knight. Still White must be a little bit better here. 8 . lLlc3 Wic6 9 .Wif3 lLle7 (9 . . . .ie6 10 . .ig5 f6 1 1 ..id2;!; Ramesh -Ferrufino, Bled 2002 looks reasonable to me) And now it is of course possible to go 1 0 .g4 .ie6 l 1 . lLle3 l:'!:d8 1 2. lLl ed5°o Hector - Bator, Sweden 2003. But White should also do well with the simple 10 .lLlxe7 .ixe7 1 1 .Wig3;l; . 5.�c3 e5 6.i.b5t �bd7

6 . . . .id7 7 . .ixd7t Wixd7 leads to a slightly inferior position after 8 . lLl de2! (8. lLlf5?! lLlxe4 9 .lLlxg7t �xg7 10 .lLlxe4 d5 1 1 ..ih6 0-0 leads to an equal position, or maybe even a forced draw. One line is 1 2 .lLlc5 Wid6 13 .�xg7 �xg7 14 .lLlxb7 �b4t 1 5 .�d2 Wixb7 Schopf - Hendrix, e-mail 1 997) . Now Black has the following possibilities:

a} 8 . . . Wi g4! ? This looks very strange, but it does win a pawn. 9 .�d3!? A


very aggressive approach that left Black with material but without development. 9 o o .1/:;\fxg2 (9 . . . ti:ic6 10 .0-0 i.e7 1 1 . ti:ig3 t according to Pengo However, maybe the position is just really bad for Black. The queen is utterly misplaced and 1 2. ti:i d5 would be a strong reply against 1 1 . . . g6.) 10 J !gl 1/:;\fxh2 1 1 .i.g5 ti:ibd7 1 2.0-0-0 1/:;\fxf2 13 .i.e3 'lMff3 14J �g3 1/:;\fh5 1 5 . ti:ib5 l'k8 16 .ti:ixa7 Eid8 17 . ti:ib5 ie7 1 8 . Eixg7 ti:ic5 19 .'lMfc4 ti:icxe4 20. ti:i c7t 'it>d7 2 1 . Eixf7 and the attack was very strong in Willemze - S. Ernst, Vienna 2003. Possibly Black can defend better at various places, but what about 9 .ti:id5! which leaves White with a small but lasting advantage after 9 . . . ti:i xd5 1 O.1/:;\fxd5 ti:i c6 1 1 . ti:ig3 1/:;\fe6 1 2.c4. Compared to the Kalashnikov, White's bad bishop has been exchanged, while Black will find it hard to exchange his bad bishop, in contrast to the Kalashnikov. 9 .ti:ig3 'lMfxdlt 1 0. 'it>xdl also looks like a preferable position for White.

b) 8 o o .i.e7 9.0-0 0-0 1 0. ti:i g3 g6 is no way to play chess. 1 1 . i.g5 ti:i c6 1 2.1/:;\f d2 'it>h8 13 . Eiadl Eiad8 14.f4 ti:ig8 was played in Peng - Alexandria, Jakarta 1 997. Now Peng gives 1 5 .f5 i.xg5 1 6.'lMfxg5 f6 17. 'IMf e3 ± as the best continuation.

c) 8 o o .h6. This seems to be the most serious move. But the fact is that White is a little better no matter what. 9.0-0 (9. ti:ig3!? with the idea of9 o o .g6 1 0. ti:i fa) 9 . . . ti:i c6. So far Paehtz - Petrenko, Plovdiv 2003. In the game I think Black could

have equalised, but after 1 O .ti:ig3 White is simply slightly better. 7.tLlf5 a6 S.J.xd7t 'ilYxd7

9.tLle3� This is a very modest decision, far from

an outright refutation of Black's opening. Usually when I meet an unusual line I do not try to refute it, unless there is no other way of gaining an advantage, or unless I think it is straightforward to refute it over the board. Here this kind of thinking explains my choice. Instead of 9. ti:i e3 White has a more aggressive alternative:

9 .i.g5 ti:ixe4 1O .ti:ixg7t i.xg7 1 1 . ti:ixe4 0-0 12 .1/:;\fxd6 f6 13 .1/:;\fxd7 i.xd7 14 .i.d2 i.c6 1 5 . ti:i d6 has been played a number of times, and this also leads to a slight advantage for White. The two bishops and the strong pawns in the centre do provide Black with some counterplay though, and I do not feel that this is the most challenging way to play with


White. The positional approach, where d5 remains weak, is more to my liking. 9 ••• YlYc6 10.YlYd3

There is an alternative that also really appeals to me. 10 . �f3!? i.e6 1 1 .0-0 i.e7 1 2. ltl f5 .txf5 13 .�xf5 �d7 14 .'�f3;!; Strautins - Gallerani, e-mail 2002. 10 •• .i.e6 1 1 .0-0 gc8

1 1 . . .i.e7 is also playable. 1 2.a4 0-0 1 3 . .td2 gfc8 14.a5 b5 1 5 .axb6 �xb6 Marinkovic - Rajkovic, Ulcinj 1 997. 1 6. ga4!? i.d7 (Untenable is 16 .. . YlYxb2? 17 .gbl gxc3 1 8 . gxb2 gxd3 1 9.cxd3±) 17. ltl cd5 ltlxd5 1 8 .ltlxd5 YlYd8 19 . 9xa6 gxa6 20. YlYxa6 and White would be laughing, as 20 . . . gxc2 2 1 . i.c3 traps the rook. 12.a4

White has won the opening batde as far as I am concerned. He has exchanged the correct pair of bishops and prevented Black's counterplay on the queenside. Soon he will advance his pawns and create lasting problems for Black on the queenside. 12 • • • J.e7 B.a5 YlYe5

1 3 . . . 0-0 14 . ltlcd5 i.d8 1 5 .c4 ltl d7 1 6.b4 i.xd5 17.exd5 �c7 1 8 . .ta3;!; Della Morte - Larrea, Vicente Lopez 2004. 14J'idl 0-0 15.J.d2 �Ud8 16.i.el J.fB 17.�a4 YlYc6 18.�b6 ge7 19.f3 �h5 20.e4� White later won. Hjartarson -

Bator, Gausdal 1 996.

This provocative move was invented by Bent Larsen who has played it from time to time, never presuming that it is very good, but to get young players out of theory. In the game fragment below his much younger opponent shows him the drawback with this kind of thinking: Bad moves often lead to bad positions!

1 .e4 e5 2.�f3 d6 3.d4 cxd4 4.�xd4 �f6 5.�c3 �bd7 6.g4!

This is the strongest reply, both statistically and logically. Black will be badly placed to play a hybrid between the English Attack and whatever he is trying to do.

6 ••• �e5?! I think this move is questionable. 6 . . . d5 does not look good once we have

a deeper look at the practical examples. 7.exd5 (7. ltlxd5!? is a serious alternative. 7 . . . ltlxd5 (7 . . . ltlxe4? 8 . ltlb5+-) 8 .exd5 �a5t (8 . . . ltl f6 does not work on


account of 9 . .tb5t .td7 10 .c4±) 9.c3 'iNxd5 10 . �gl and White is probably slightly better here. One line is 10 . . . e5 1 1 .'iNe2 .tc5 12 .,tg2 VMd6 1 3 .ltlf5 'iNf6 14.,te3;1;) 7 . . . ltl b6 8.,tb5t .td7 9.d6 a6 as played in Feher - Gross, Budapest 1 998. (9 . . . exd6 1 0. ltl f5 .txb5 1 1 . ltlxb5 d5 12 . .tf4± was something White quickly made a full point of in van Asperen -Barks, e-mail 1 998) Here White played the innocuous 1 0.,te2?! and the game was eventually drawn after the strongest reply 10 . . . e5! . But White had a much stronger continuation with 10 .g5!

when I cannot find a playable move for Black! 1 0 . . . .txb5 ( l 0 . . . axb5 l 1 .gxf6 gxf6 1 2. ltl dxb5 either transposes or gives Black problems after 1 2 . . . exd6 1 3 .ltlxd6t ,txd6 14 .VMxd6±) 1 1 . ltl dxb5 axb5 1 2.gxf6 gxf6 1 3 .ltlxb5 exd6 14 .VMe2t .te7 15 . .tf4± and Black's position is a complete wreck. I am sure a stronger player than I would go all the way and

say that White is winning . . . 6 . . . h6 i s probably Black's best move

here. I would imagine that 7.h4!? is a good reply. 7.,te3 a6 (7 . . . h5? is punished with 8.g5 ltlg4 9.g6! fxg6 1 0. ltl e6 ltlxe3 1 1 .fxe3 VMa5. This was all played in Gaggiottini - Beggi, Italy 1 995 . Now the computer quickly finds the winning move: 12 .VMd5!! VMxd5 1 3 .ltlxd5 l'!b8 14. ltl dc7t 'it>f7 1 5 .,tc4+-) 8 .h4 g6 9 .VMe2 h5 1 0.gxh5 ltlxh5 1 1 .0-0-0 .tg7 12 . �gl ltlc5 1 3 .f4;1; with a good attacking position for White in Fogarasi - Bilek, Zalaegerszeg 1 992. 7.a g6

This is one way to play the position for Black. But really he has no easy options anymore. 7 . . . e6 8 . .te3 a6 9 .'iNd2 b5 1 0.a3 ! ? Strictly speaking this is not necessary. 1 0 . . . .tb7 1 1 .g5 ltlfd7 1 2.0-0-o �c8 1 3 .'it>b 1 .te7 14.h4;1; . These kind of attacking positions do not usually allow a player to waste time, as Black has done here with the artificial knight manoeuvre. White won in Barczay - Ciocaltea, Varna 1 967, though Black is still in the game at this moment. 8.!e3 !d7

8 . . . a6 9 .VMd2 b5 1 0.0-0-0;1; is pleasant for White. Here Black is not really ready to face ideas such as ltl d5 followed by ltl c6, or just the basic g5 and ltlc3-e2-g3 . Black has a problematic choice to make. 1 0 . . . .tb7?! is, however, not the way to go . After 1 1 . ltl dxb5! White had just won a pawn in Shevelevich - Makarov, corr.


1 985 , based on l 1 . . .axb5 12 .!xc5 dxc5? 1 3 .!xb5t and White wins. 9.h4 a6 10.h5

I quite like White's flexible play in this game. Already here he must have been thinking of l 1 .b4 and the later !c4 and 'l&d3 . However 10 .'I&d2!? , with the idea of 1 1 .0-0-0, would also have given White the better game. 10 •.• b5

1 l .b4! This move is rather surprising. Usually

White would not commit to an offensive on both sides of the board. Here it is fully justified. 1 l ••• �a4

It is hard to argue with this move, even though Black will be worse after it. The computer wants to play rough with 1 1 . . . e5, but White has a refutation in 12 .g5! lDxh5 1 3 .bxc5 exd4 14 .!xd4 E:g8 1 5 . lD d5± , based on ideas with E:xh5, and 15 . . . dxc5 1 6.!e5 and White wins.

12.lDxa4 bxa4 13.hxg6 fxg6 14.,ic4 tvc8

14 . . . e5 1 5 . lD e6 �c8 16. lDxfB �xc4 17. liJ xd7 liJxd7 1 8 .�xd6 �c3t 19 . �f2 �xc2t 20. �g3± was no alternative either. Black's position is simply bad. 15.tvd3 i.g7 16.0-0-0 a5 17.b5 gbS IS.a3±

White won an exciting game and later the Danish Championship in a play-off between the same two players. Mortensen - Larsen, Aalborg 1 989.

5 . . . J.d7

This variation is the best of the three 5th move alternatives, and is usually attributed to the Byelorussian grandmaster Kupreichik.

1 .e4 c5 2.�f3 d6 3.d4 cxd4 4.�xd4 �f6 5.�c3 i.d7 6.f3!

I think the English Attack is the most natural reaction to this move. Black now has the option to go into the unusual lines of the Dragon, or play a position where the bishop is a little strangely placed on d7.

6. !g5 will most often transpose to the Richter-Rauzer after 6 . . . lDc6. This is, of course, something White can choose to play, but I like to recommend the best move in the position.

6.!e2 is another way to play the position, but is not in line with our repertoire.


6 .. . a6 I am not convinced this is the best plan.

The alternatives are: 6 . . . e5 7. tLlb3 a5 8 .a4 !i.e7 9 .!i.e3 tLla6

10.!i.b5 ;!; Topalov - Ivanchuk, Monte Carlo 1 998.

6 . . . tLlc6 7.!i.e3 g6 (7 . . . a6 8 .1Mfd2 �c8!? is an interesting idea that cannot be correctly evaluated before it has been played between stronger players than it has currently.) 8 .1Mf d2 transposes to the Dragon variation, see page 43. In this way, choosing 6.8 is maybe just a question of which transposition to allow. 7 • .te3 e6 8.'i!Yd2 b5 9.0-0-0�

White is a little better after the opening. 9 . • • h5 10 • .td3 .te7 l l .g3!?

A slow but dangerous plan. Black needs to react in the centre as in the game, or he will be in trouble. l l .•• 'i!Yc8 12.h3 e5!

1 2 . . . tLlc6 1 3 .g4 with an advantage for

White. 13.tLlf5 ixf5 14.exf5 d5 15.f4 d4

1 5 . . . e4? 1 6.!i.e2 !i.b4 17.!i.d4 tLl bd7 1 8 .g4± and Black will soon find himself with very serious problems. 16. 'i!Y g2 'i!Y c6

1 6 . . . tLl bd7 17.fXe5 dxc3 1 8 .exf6 cxb2t 1 9 .�bl tLlxf6 20.!i.d4 0-0 2 1 .g4± leads to a position where White has good attacking prospects. 17.tLle4! dxe3? This simply loses the exchange without compensation. After 17 . . . tLl d5 18 .!i.f2. f6 1 9 .fXe5 fXe5 ;!; Black is worse, but there is no clear way for White to break down Black's defences. 18.tLlxf6t .txf6 19 . .te4 'i!Yc7 20.J.xas White is winning, and won in another 20 moves. Movsesian - Markovic, Jahorina 2003.

286

Index of variations

Najdorf 1 .e4 cS M:J6 d6 3.d4 ad4 4,lilxd4 �f6 S.�c3 a6 6 • .tgS 6. . .�bd7 7.f4 Wb6 10 6 ... e6 7.£4 7 . . . � c6 S.e5 h6 9 . .th4 g5 14 (9 . . . �xd4 12, 9 . . . dxe5 12) 7 . . . We7 17 7 . . . b5 S .e5 dxe5 9.fxe5 We7 1 0.exf6 We5t 1 1 ..te2 Wxg5 1 2.0-0 l":!a7 18 ( l2 . . . We5 20) 7 . . . .te7 24 (7 . . . .td724) 7 . . . h6?! 31 7 . . . Wb6 S.Wd2 Wxb2 (S . . . �e6 36) 9.�b3 36 (9.l":!b I 36) 7 ... �bd7 8.W6 S . . . Wa5 24 8 ... Dc7 9.CJ-O.-O bS 1 0 . .ixb5 29 1 0.e5 29 10 . hf6 29 10 • .td3 .tb7 1 1 . ghel I l . . .Wb6 27 1 1 ... .te7 32

Dragon 1 .e4 cS 2.�6 d6 3.d4 ad4 4.�xd4 �f6 S.�c3 g6 6 • .td .tg7 7.6 7 . . . a6 S.Wd2 �bd7 45 7 . . . 0-0 S .Wd2 d5?! 45 7 ... �c6 8.Wd2 S . . . .td7 9.0-0-0 l":!eS 47 8 ... 0-0 9.CJ-O.-O 9 . . . .td7 48 9 . . . �xd4 1 0.ixd4 ie6 61 ( l0 . . . Wa5 55) 9 . . . .te6 60 9 ... dS 10.exdS �xdS 1 1 .�xc6 bxc6 12..td4 12 . . . e5 13 . .te5 .te6 ( 1 3 .. J:l:eS 70) 14.�e4 l":!eS 73 ( l4 . . . l":!bS 73 14 . . . We7 73) 1 2 . . . �xc3 85 12 .. ..txd4 13:f!Yxd4 1 3 . . . Wb6 81 13 ... Y!Yc7 81

Sveshnikov 1 .e4 cS 2.�6 �c6 3.d4 ad4 4.�xd4 �f6 S.�c3 eS 6.�dbS d6 7..tg5 a6 8.�a3 S . . . .te6 91 8 ... bS 9.hf6 gxf6 10.�d5 1O . . . .tg7 1 1 .c3 f5 tranposes. 10 ... 5 1 1 .c3 .tg7 12.exf5 hf5 13.�c2 13 . . . .te6 107 13 ... 0-0 14.�ce3 .te6 14 . . . .tg6!? 95 14 .. ..te6 IS..td3 5 16.0-0 1 6 . . . lt>hS!? 95 1 6 . . . e4 105 16 ... ga7 17.a4 �e7 18.�xe7t gxe7 19.axbS axbS 20 . .txf5 97 20.ic2 20 . .bbS 99

Classical 1 .e4 cS 2.�6 d6 3.d4 ad4 4.�xd4 �f6 S.�c3 �c6 6.ig5 6.. .id7 112 (6 . . . Wb6 112, 6 . . . g6 1 12, 6 . . . Wa5 112) 6 ... e6 7.Y!Yd2 7 . . . a6 S.o-o-O .td7 126 (S . . . h6 120) 7 . . . Wb6 132 7 ... .te7 8.CJ-O.-O 8. . .0-0 137 8 ... a6 9.£4 �xd4 10.Y!Yxd4 133

Kan 1.e4 cS 2.�6 e6 3.d4 ad4 4.�xd4 a6 S.�c3 5 . . . We7 6 . .td3 �f6 (6 . . . .te5 155) 7.0-0 ie5 S.�b3 ia7 157 (7 . . . ie7 153) S ... bS 6 . .td3 6 . . . Wb6 1 60 6 . . . . ib7 7.0-0 Wb6 1 64 (7 . . . We7 168) 6 . . . d6 166 6 . . . ie5 7.�b3 L7 (7 . . . ie7 170)

Taimanov 1 .e4 cS 2.�6 e6 3.d4 ad4 4.�xd4 �c6 S.�c3 5 . . . a6 6.ie3 �f6 185 (6. . .�ge7 185) S ... Y!Yc7 6 • .td a6 7..td3


7 . . . b5 171 7 ... �f6 8.0-0 8 . . . .id6 173 8 . . . h5?! 176 8. . .�xd4 178 8 . . .�e5 180 8 . . . d6 182

Accelerated Dragon 1 .e4 c5 2.�f3 �c6 3.d4 cxcI4 4.�xd4 g6 5.c4 .ig7 6.�c2 �f6 7.�c3 0-0 8.ie2 d6 9.0-0 9 . . . id7 1 0 . .ie3 197 9 ... �d7 lO.idl 1 0 . . . a5 194 10 ... �c5 1 1 .b4 �e6 192 ( 1 1 . . .ixc3 190)

Scheveningen 1.e4 c5 2.�f3 d6 3.d4 cxcI4 4.�xd4 �f6 5.�c3 e6 6.g4 6 . . . d5 199 6 . . . e5 200 6 . . . .ie7 200 6 . . . a6 7.g5 �fd7 8.h4 b5 9.a3 .ib7 10 . .ie3 �e6 205 ( 10 . . . �b6 201, 1 0 . . . �e5 203, 1 0 . . . .ie7 204) 6. . .�e6 206 6 ... h6 7.h4 7 . . . a6 210 7 . . . .ie7 210 7 ... �c6 8.gg1 8 ... h5 213 (8 . . . d5 21 1)

Kalashnikov 1 .e4 e5 2.�f3 �c6 3.d4 cxcI4 4.�xd4 e5 5.�b5 d6 6.� lc3 a6 7.�&3 b5 8.�dS 8. . .�ge7 219 (1l . . . ie7 2 1 9, 8 .. .l:!b8 219) 8 . . .�ee7 222 8 ... �f6 9.c4 b4 223 (9 . . . �d4 223)

The Four Knights 1 .e4 c5 2.�f3 �c6 3.d4 cxcI4 4.�xd4 �f6 5.�c3 e6 6.�db5 6 . . . .ie5 234 6 ... ib4 7.&3 ixc3t 8.�xc3 dS 9.exd5 9 . . . �xd5 228

9 . • . exd5 10.id3 0-0 1 1 .0-0 l 1 . . .h6 229 l 1 . ..ig4 229 1 1 ... d4 231

The Pin variation 1 .e4 c5 2.�f3 e6 3.d4 cxcI4 4.�xd4 �f6 5.�c3 ib4 6.e5 6 . . . �e4?! 236 6 ... �d5 7 .idl �xc3 8.bxc3 ie7 8 . . .ia5 239 8 . . . .ifS 245 9.Y!Yg4 9 ... l!?flI 246 9 . . . 0-0?! 243 9 . . . g6 243 9 . . . g5 243

The Nimzowitsch Variation 1 .e4 c5 2.�f3 �f6 3.�c3 3 . . . d5 249 3. . .�e6 4.d4 d5 252 3 ... e6 4.e5 �dS 5.�xd5 exdS 6.d4 6 . . . d6 255 6...�c6 7.dxc5 ixc5 8.Y!Yxd5 d6 260 8 . . . YHb6 256

Minor Lines 1 .e4 c5 2.�f3 2 . . . b6 264 2 . . . a6 266 2 . . . g6 3.d4 .ig7 269 2 . . . e6 3.d4 cxd4 4.�xd4 YHb6 273 (4. . .ie5 273) 2 ••• �c6 3.d4 cxcI4 4.�xd4 4. . .�f6 5.�e3 e5 6.ltJdb5 h6 268 4 . . . YHb6 271 4 . . . e5 5.ltJb5 a6 276 4 ... e6 5.1tJc3 d6 277

5th Move alternatives l .e4 e5 2.1tJf3 d6 3.d4 cxd4 4.ltJxd4 1tJf6 280 4 . . . e5 280 5.1tJc3 id7 284 5 . . . e5 280 5 . . . ltJbd7 282

288

Index of games :

Game No Players page no 33. Smirin - Markovski 1 66 1 . Stripunsky - Granda Zuniga 1 0 3 4 . Ponomariov - Gallagher 1 68 2. Adams - Anand 12 35 . Hector - Lindberg 17 1 3 . Shabalov - Browne 14 36 . Hector - Pogorelov 1 73 4. Khalifman - Lautier 1 6 3 7 . Parligras - Miladinovis 1 76 5. Leko - Ghaem Maghami 1 8 3 8 . Almasi - Piket 178 6. Wosch - Nordin 20 39. Ponomariov - Sadler 1 80 7. Lautier - Efimenko 24 40. Yagupov - Khusnullin 1 82 8. Shabalov - Gelfand 27 4 1 . Hector - C. Hansen 1 8 5 9. Timman - Gelfand 28 42. Svidler - Tiviakov 188 10 . Kotronias - Lesiege 3 1 43. Aronian - Vorobiov 1 92 1 1 . Short - Kasparov 35 44. Bologan - Motylev 1 94 12 . Palac - Lalic 44 45. Gulko - P. H. Nielsen 1 97 13 . Golubev - Poliantsev 55 46 . Renet - Summermater 205 14. van der Wiel - Sax 60 47. TImmermans - de Jonghe 207 1 5 . Ehlvest - Marin 69 48. Karpov - Spassky 2 1 1 16 . Balinov - Velickovic 8 1 49. Morovic Fernandez - Veingold 2 1 4 1 7 . Rowson - Mah 85 50. Motylev - Shariyazdanov 2 1 9 1 8 . Yemelin - Kharlov 9 1 5 1 . Delchev - Brumen 222 19 . Zeleic - Zelenika 94 52. Anand - Shirov 223 20. Hector - Carlsen 99 53. Timoshchenko - Chernov 228 2 1 . Rivas Romero - Sarlat 105 54. Karjakin - Raetsky 23 1 22. Jenni - Avrukh 107 55. Hansen - Kristensen 243 23. Kotronias - Schwartz 1 1 2 56. Varga - Horvath 245 24. Balashov - M. Makarov 120 57. Boriss - Bartsch 249 25. Lastin - Spragetr 126 58 . Ismagambetor - Palit 252 26. Dolmatov - M. Makarov 132 59. Braun - Choroba 255 27. Iordachescu - Campos Moreno 137 60. Hlavac - Bazant 260 28. Acs - Fancy 1 52 6 1 . lhorhallsson - Balinov 264 29. van der Wiel - Nijboer 1 5 5 62. Shirov - Kurajica 266 30. Predojevic - Fogarasi 1 57 63. Movsesian - Kozul 271 3 1 . Adams - Kasimdzhanov 159 64. Yu Shaoteng - Zaw Win Lay 273 32. Svidler - Milov 164 65. Senff - Schlosser 277

aagaard, jacob & shaw, john - experts vs. the sicilian (2004, 1st ed)

Documents