player movement patterns and game activities in the australian football league

rRAINING ISSUES

Player movement patterns and game activities in the Australian Football League

B Dawson 1,3, R Hopkinson 1, B Appleby 2'3, G Stewart 3 & C Roberts 4

1School of Human Movement and Exercise Science, The University of Western Australia. 2Western Australian Institute of Sport. 3West Coast Eagles Football Club. 4Western Australia Football

Commission.

In the Australian Football League (AFL), specifc game movements and activities have not been studied since the 1970s and 1980s and the game is now much faster than it was 20- 30 years ago 4. Using lapsed-time video analysis, AFL players from five different positions (full forward/full back, centre half forward/centre half back, small forward/small back, mid fielders and ruckmen) had their movement patterns (stand, walk, jog, fast-run, sprint, change of direction) and game activities (possessions, ruck duels, ground ball contests, shepherds, spoils, bumps and tackles) recorded in two games each in the 2000 season. A descriptive analysis of the results was undertaken. The main findings were: full forward/full back were most different from the other positions, as they were seen to stand more and jog and fast-run less; ruckmen and midfielders were involved in more game activities than the other positions; for all positions, there were more than 150 high intensity movements (fast- run plus sprint) in the game, but these accounted for only 4-6% of total movement time; virtually all of the high intensity movements lasted for <6 secs; more than half of all sprints involved at least one change of direction, mostly within the 0-90 ° arc (left or right) and all ground ball contests took <6 secs, with midfielders having 2-3 times as many as the other positions. Improvements in specific pre-season and in-season training practices for different positions should be possible after careful interpretation of these findings.

(J Sci Med Sport 2004;7:3:278-291)

Introduction Of all the major football codes, Australian Football appears the most 'under researched', at least with respect to identifying the physical and physiological d e m a n d s t h a t a r e p l a c e d u p o n p l a y e r s d u r i n g a game . A la rge -sca le , coin- p r e h e n s i v e s t u d y of exac t ly w h a t m o v e m e n t s a n d t a s k s a p l a y e r m u s t p e r f o r m in a m a t c h h a s neve r r ea l ly b e e n u n d e r t a k e n . Therefore , knowledge of speci f ic game d e m a n d s , on w h i c h p r e - s e a s o n a n d i n - s e a s o n t r a i n i n g p r o g r a m s s h o u l d be b a s e d , h a s n o t b e e n r ead i l y ava i lab le . Seve ra l s m a l l - s c a l e s t u d i e s on m o v e m e n t p a t t e r n s in A u s t r a l i a n footba l l were p e r f o r m e d in t h e 1970s a n d 1980seg 1,2,a. However, a s t h e s e s t u d i e s were p e r f o r m e d b e t w e e n 15 a n d 30 y e a r s ago a n d h a d seve ra l s h o r t c o m i n g s , t he i r r e l evance to t he m o d e r n g a m e is now se r ious ly l imi ted . It is n o w wide ly a c c e p t e d t h a t t he g a m e h a s b e c o m e fa s t e r s ince t h e s e s t u d i e s were c o n d u c t e d 4, a s t he g a m e h a s evolved f rom semi - p ro f e s s iona l S t a t e League c o m p e t i t i o n s in to t he ful ly p r o f e s s i o n a l A u s t r a l i a n Foo tba l l League (AFL) n a t i o n a l compe t i t ion . Therefore , i t is dif f icul t to a c c e p t t he r e s u l t s of the c u r r e n t l y - a v a i l a b l e l i t e r a t u r e a s typ ica l g a m e r e s p o n s e s for t he AFL today , p a r t i c u l a r l y a s t he s a m e p l a y e r s have n e v e r b e e n a n a l y s e d in m o r e t h a n one game a n d in s o m e c a s e s have b e e n m o n i t o r e d for one a n d two q u a r t e r s only.

278

Player movement patterns and game activities...

Further, very little knowledge on game activities (ie, player, ball and body contacts) is available, as these studies provide virtually no information on these aspects . Recently, Appleby and Dawson 5 presented da ta regarding the frequency, type and location of mark ing contests, bounce-downs and throw-ins in AFL matches , bu t individual player activities were not recorded. Therefore, it is not known whether common training activities adequately ma tch game demands . Also, a comprehensive analysis of the several positional roles within the t eam in regard to their specific game requi rements is lacking. It is apparen t then tha t the majority of existing Austral ian Football research knowledge, from which the development of t e am and individual player conditioning and training p rograms should flow, is very out-dated. There is an obvious need for fur ther research into Austral ian Football, in par t icular a comprehensive s tudy of exactly wha t movements and activities a player m u s t perform in a match. Therefore, the purpose of this s tudy was to document the specific game demands (movement pa t t e rns and ma tch activities) of different posit ions in AFL ma tches so tha t cur ren t information is available on which to develop appropr ia te training programs. A companion s tudy was also performed to investigate how closely the documented game demands were replicated during in-season training sessions for the same players. These resul ts are reported in a companion paper.

Methods Subjects One player from either the West Coast Eagles or Fremant le Dockers was videotaped for the whole game in every AFL match played in Perth in the 2000 season (22 games in all). Each player was videotaped (Sony DSR-200P camcorder) by a professional c a m e r a m a n in two complete games (barl4mg injury or interchange), from an elevated position in the g rands tand suitable for obtaining an unin te r rupted view of the oval. Prior intention to film on a given day was not given to rely player, ensur ing tha t da ta collection was "single blind". However, informed consent regarding the possibility of being filmed in games was received from the players of bo th t eams prior to the commencemen t of the study.

Video Procedures If a player selected to be videotaped on a given day was injured or interchanged during a game, the player who replaced him was then filmed, as long as he took up the same position. That is, the same position (rather t han player) was filmed within a game. Similarly, if a player being filmed was moved to a different position within a game, this information was conveyed to the c a m e r a m a n and a new player competing in the originally selected position was then followed. The n u m b e r of occasions this occurred was recorded. The playing conditions (dry or wet) were recorded for each player during games. However, the conditions had no influence on player filming.

Players were selected from the following positional or player types: a) Full Forward/Ful l Back (FF/FB) - more 'static ' key position players (n= 1

forward and 1 back) b) Centre Half Forward /Cent re Half Back ( C H F / C H B ) - more mobile tall

forward or back players (n= 1 forward and 1 back)

279


c) Forward Pocket /Half Fo rward /Back Pocket /Hal f Back - more mobile small forward or back (SF/B) (n= 1 forward and 1 back)

d) R u c k m a n - likely to contest mos t bounce-downs / th row- ins and doesn ' t normal ly play a 'set ' position (R) (n= 2)

e) Rover /Ruck R o v e r ~ C e n t r e ~ W i n g - mid-fielders (n= 3) Reliability and consis tency checks were made by having all testers analyse a

selected game on at least two occasions, after first refining their m e a s u r e m e n t techniques and definitions of game movements and activities by viewing a practice ma tch on at least two occasions. To fur ther minimise any in ter - tes ter effects on the data, one tes ter each was assigned to analyse respectively: a) all movement pa t t e rns (standing, walking, jogging, fast running and

sprinting), except lateral movements b) all lateral movements (left or right changes of direction in sprints) c) all game activities, except ground ball contests d) all ground ball contests (number and dura t ion of each).

Therefore, only one tes ter was responsible for analysing all of the ma tch videotapes for each of these variables. The total n u m b e r and, where appropriate, the total t ime of game movements and activities were analysed. Test- re test reliability for each tester was assessed by applying t test, Pearson Correlational analysis and Technical Error of Measurement Calculat ions (TEM) to the da ta obtained from the first and second trials of the game selected for reliability assessment . In summary , t tes t values for all of the variables were non-significant, and Pearson correlation coefficients were r= 0.96 or greater. Generally, the TEM values were fewer t han 4% for the total n u m b e r of movements and activities and 7-11% for total times.

The videotapes were reviewed and analysed for the following movements and actions tha t the player made during the games filmed:

Movement Patterns The following movements were categorised, based on the descript ions used by Mayhew and Wenger 6 and Durward et al 7. i) Standing (no movement in any direction) ii) Walking (in any direction) iii) Jogging (slow, easy, unhur r ied running in any direction) iv) Fast running (more purposeful running bu t not nea r maximal speed) v) Sprinting (running at top speed or very close to it)

(standing, walking and jogging were considered to be "low intensity" efforts, while fas t - running and sprint ing were considered to be "high intensity" efforts).

vi) Lateral movements (left or right changes of direction in spr ints and recorded as 0-45 ° , 46-90 ° , 91-135 ° or 136-180 ° turns)

vii) Work : recovery ratio (time between high intensi ty [fast running or sprinting] efforts)

The t ime spent in each movemen t pa t t e rn was recorded, along with a specific b reakdown of the durat ion and frequency of fast running and sprint ing repeti t ions (eg, 25 x 0-3-sec fast runs, 10 x 6-9-sec fas t - runs , 15 x 0-3-sec sprints, 5 x 3-6-sec sprints, etc). To est imate the dis tance covered when

280


walking, jogging, fast-running and sprinting, the players were filmed at a training session (after their two games taped for analysis were completed) when moving at each of these four speeds over a 20-metre grid. The players were instructed to move at the speeds they would typically use for each of these movement pat terns in a game. Two trials for each movement pattern were filmed. Values for steps per second and metres per step were then obtained (from an average of the two trials), and applied to the total time spent walking, jogging, fast-running and sprinting to provide an estimate of the total metres covered in each movement pattern. An example of this procedure is included below:

Example (Walking) Steps/Second: = Number of steps for 20 m/Time for walking 20 m = 22 /12 .4 secs = 1.77 Metres/Step: = 20 me t r e s /number of steps = 20 m / 2 2 = 0.91 Total Metres Covered in Walking = Metres/Step x Total Time spent Walking x Steps /Second = 0.91 x 3351 secs x 1.77 = 5397 m

Game Activities The following activities were recorded and categorised: i) Ball possessions

Kicks (contested and uncontested) Marks (overhead or chest, contested or uncontested, taken or missed) Handballs (contested or uncontested)

ii) Ruck Duels (bounce-downs and throw-ins, contested and uncontested) iii) Ground Ball Contests (any situation, except a marking or ruck contest,

which involved competing directly with an opponent to at tempt to win the ball when 'in dispute')

iv) Shepherds, bumps, smothers, spoils and tackles (both given and received) v) Going to ground (either diving or being knocked/pushed) vi) Other (such as knock-ons in general play)

For marks, kicks, handballs and ruck duels, a 'contested' or 'uncontested ' involvement was recorded. A 'contested' involvement meant that the player was under direct physical pressure from an opponent. That is, there was already body contact between the players or the opponent was within one metre and therefore close enough to effect imminent body contact.

The number of each of these games activities was recorded and the ground ball contests were also timed so that a specific description of these actions could be provided: for example, ground ball contests, 15 x 1-2 secs, 12 x 2-4 secs, etc. Timing of the ground ball contests was stopped when a break in play occurred (eg, free kicks or bounce-downs/throw-ins) or when the ball was moved from the immediate area of the filmed player, such that his effort noticeably slackened.

Data Analysis To facilitate data analysis a customised touch pad was constructed (Prowess Systems; Melbourne) so that game movements and activities could be time- coded a n d / o r recorded by pressing the appropriate key as the videotape was viewed. The information was then logged into a database from which a report

281


file (Crystal reports writer version 8, Seagate, Melbourne) was produced, which was then downloaded, so tha t averages for each player and then in turn, each posit ional category, could be produced.

Although each player was filmed in two games, due to the small total n u m b e r of players (n= 2-3) filmed for each positional type, no inferential s tat ist ics were per formed on the data. Only descriptive statist ics (mean _+SD and range) were calculated, as the intent and purpose of the s tudy was only to record accurate ly the movemen t pa t t e rns and activities performed by the players in games. For ease of presenta t ion mos t num ber s were rounded up to the neares t whole n u m b e r and 0.5 resul ts were rounded up.

Results The resul ts included here are the averages of the values recorded within each position category (n= 4 or 6 games). Due to a camera malfunction, one r u c k m a n was only filmed in 1.5 games; therefore ma tch totals for this positional category are the average of only three games instead of four. Games were a lmost all played in dry conditions; only two were played in wet wea ther (one was for the midfielder position and the other for CHF/CHB position). The resul ts include s i tuat ions where a rep lacement player for the same position (due to injury or interchange) was used in par t s of games (on 13 separa te occasions). In all cases, this was for less t han one quar ter of the total game.

Movement Patterns The average ma tch totals for the n u m b e r of occasions tha t the players engaged in standing, walking, jogging, fas t - running and sprinting, as well as the total t ime spent in each of these movements are shown in Table 1. As percentages of the total, the FF/FB spent 23% standing, while the other posit ions ranged from 11-16%. Walking comprised 47-54% for all positions, with FF/FB and SF/B being the greatest . Jogging was approximate ly 35% for midfielders, r uckmen and CHF/CHB, reducing to 26% for SF /B and only 19% for FF/FB. Fas t - running was 6% for the midfielders, 4-5% for the ruckmen, CHF/CHB and SF/B bu t only 3% for the FF/FB, while sprint ing was 1% for FF/FB and SF/B, bu t only 0.6% for the other positions.

Table 1 also includes an es t imate of the average dis tances covered by each of the positional types when walking, jogging, fas t - running and sprinting. The midfielders were found to cover the greates t es t imated average distance, being slightly less t han 17,000 m in total and the FF/FB the least, being j u s t over 13,600 m. Ruckmen covered nearly 15,400 m, and the CHF/CHB and SF/B both were es t imated to cover j u s t over 16,000 m in a game.

If only f a s t - r u n n i n g and spr in t ing are cons idered , these combined movements only represen t a small percentage (4.4% in FF/FB, to 6.3% in midfielders) of the total movement time. Therefore, m u c h of the total game t ime is spent standing, walking or jogging. However, as shown in Table 2, these high in tens i ty m o v e m e n t s toge ther average be tween 158 (FF/FB) and 208 (midfielders) in a game for the five positional types. Sprints accounted for only 10-19% of the total n u m b e r of high intensi ty movements , and were generally preceded by fast-running. The FF/FB and SF/B recorded the mos t spr ints (30) in a game, while the taller players ( ruckmen and CHF/CHB) had the fewest (17- 18). For fast running, midfielders had the grea tes t n u m b e r (185) and FF/FB the fewest (128), which ma tches the da ta presented in Table 1, showing tha t

182

Player movemen t patterns and game activit ies...

Position/Movement No. Time (sees) Time (rain) Estimated Distance (metres)

FF/FB - Standing 180 .+ 36 1738 -+ 493 30.0 (148-220) (1271-2220)

- Walking 369 -+ 42 4022 +_ 322 67.0 6217 (322-410) (3554-4286) (5868-7138)

- Jogging 312 + 135 1400 -+ 573 23.3 5,270 (187-435) (863-2140) (2847-8569)

- Fast run 128 _+ 55 253 + 69 4.21 532 (65-174) (156-312) (949-1891)

- Spr int 30 + 9 77 + 19 1.3 595 (20-40) (60-101) (502-733)

13614 (10761-16874)

M i d f i e l d - Standing 143 _+ 18 768 -+ 120 12.8 (119-165) (640-945)

- Walking 381 + 34 3451 + 495 57.5 5442 (346-429) (2875-3919) (4167-6887)

- Jogg ing 434 + 32 2484 + 415 41.4 8642 (382-468) (2018-3202) (5602-12776)

- Fast Run 186 .+ 31 401 -- 183 6.7 2532 ( I 32-223) (193-644) (1285-4020)

- Spr int 24 .+12 49 -+ 21 0.8 359 (7-41) (19-68) (140-503)

16976 (14330-20139)

Rucl( - Standing 154 _+ 29 1052 _+ 278 17.5 (122-178) (789-1342)

- Walking 407 .+ 35 3547 ___ 157 59.1 4775 (367-433) (3418-3722) (4509-5116)

- Jogg ing 462 _+ 54 2641 .+ 587 44.0 8708 (418-522) (2199-3307) (7314-10004)

- Fast Run 162 + 24 311 _+ 82 5.2 1628 (147-189) (239-400) (1448-1977)

- Spr int 17 _+ 11 43 + 29 0.7 283 (5-26) (12-69) (86-454)

15393 (14198-17133)

Smal l F /B - Standing 210 + 60 1249 -+705 20.8 (149-279) (614-2190)

- Walking 397 + 57 3974 + 839 66.2 7029 (313-439) (3222-4912) (5270-9146)

- Jogging 379 -+ 103 1935 -+ 552 32.2 6188 (242-469) (1158-2454) (3553-8068)

- Fast Run 164 + 44 354 -+ 121 5.9 2342 (109-208) (217-487) (1701-2890)

- Spr int 31 +-- 8 79 + 36 1.3 719 (22-42) (38-122) (270-1221)

16278 (14471-18801)

CHF/CHB - Standing 161 .+ 37 789 _+ 316 13.2 (124-203) (439-1085)

- Walking 413 + 56 3505 .+ 306 58.4 5560 (346-482) (3220-3925) (5130-6255)

- Jogg ing 407 -+ 62 2443 +- 306 40.7 8266 (364-499) (2061-2797) (6853-9160)

- FaSt Run 147 .+ 24 326 __ 29 5.4 1902 (128-181) (289-356) (1653-2082)

- Sprint 18 _+ 11 41 _+ 26 0.7 278 (9-34) (20-73) (138-486)

16005 (15102-16537)

FF = full forward, FB = full back, F = forward, B = back, CHF = centre half forward, CriB = centre half back, No. = number of occasions, Time is recorded in seconds unless otherwise specified

Table 1: Movement patterns in games for different positions. Means (+_SD) and ranges (in brackets) are presented.

283

Player movement patterns and game activities,,,

+ i ~ 0 6 0 6 ~ .= . ~ E,, '..+-.- ~ s ~ o

".~_.. ' ~ ~ E:

+1~ +1";', +l,:b +16 ~

z . ~ +,~. +,~) 0 6 o 6 ~ .= ~ .

~ ~.,.,- ~.~ ~

+ l I.,;'g

,:: + ~ o ~',-- "-4> ~ i o- ,i +

i I +'~ +'~' +,~ :,~ .~ ~,,,~ ~ ~,..~,..~o . , - , o "-.,-- b ~'~ <"<>'> ~ ~, "-<~ <"'- + , - , - '~" 0")

LL " r

+,~ +,e +,~ +,7:> ~ / ~ ~ ~.~. @ .-o

= = +l";', + l ~ + l ~ +i o ~ ~ c, , , , ; ~ A

- = ~- _ , ~ =m _ ,

~. +i 'T, + i '~ , +i'T +i'T ~ ~ ,~'~ o , - - ~ & ..~

+I ~ LL

, " -~ .'-.,<" ~ "- ~ @4> 6 ~ 6 , -4> .

0 '-i,-

o o ~ ~ l u.. o (s o~ ~ / " ~ ~ ~ ~ ~ ~ "

d> ~ <s <~ ~ / ~ -~ ,, ,.,- ~ U LL

284


midfielders had the highest amount of time spent in fast-running, and FF/FB the lowest. Table 2 also shows that, for all position types, approximately 80- 85% of fast-runs and sprints in a game lasted between 0-3 secs, and 94-100% were fewer than 6 secs in duration.

Data recorded on the changes of direction during sprints showed that apart from the ruckmen (only five), on average the other positions made 15-23 changes in direction when sprinting, meaning that more than 50% of all sprints were not in a straight line. (If more than one change of direction was made in the same sprint, each one was recorded as a separate change). An approximately equal number of changes were made to the left and right and 80% or more of the total number of changes of direction were in the arc of 0- 90 ° for all positional types.

Table 3 reports the time between high intensity movements (fast-running and sprinting) in games for all the positional types. The frequencies of the ranges of times show that almost half of all times between high intensity efforts fall between 0-20 s (for all positions, this range accounted for 40-48% of the total), and approximately 70% fall between 0-60 s (67-77% for all positions).

Game Activities In Table 4 it can be seen that marks taken were similar between all the positions (3-6 marks per game). These were split evenly between overhead and chest marks and contested and uncontested marks. For marks missed, the taller players (FF/FB, CHF/CHB and ruckmen) recorded 12-17 per game, while for the smaller players (midfield and small F/B) it was only 3-5 per game, demonstrat ing that, at least for the taller players, many more marks are missed in a game than are taken. The missed marks are almost all overhead contested marks. With regard to kicks and handballs, for all positions the majority of kicks were uncontested, while most handballs (except for FF/FB) were contested.

Table 5 shows that, on average, the ruckmen had 67 (49 bounce-downs and 18 throw-ins) ruck duels per game, and that the other (especially taller) positions had very little involvement in this game action. For team involvements (tackles, shepherds, smothers and bumps), the midfielders and ruckmen tended to record higher average numbers than the other positions, reflecting a greater involvement in the play. These two positional types also went to ground a greater number of times (21-23) than did the FF/FB, CHF/CHB and SF/B (8- 11). Spoils were higher in the taller (FF/FB, CHF/CHB and ruckmen) players, indicating their greater involvement in marking contests than the smaller (midfield and SF/B) players.

Table 6 presents the total game activities for the different positions and the calculated average frequency (per second) of involvement in game actions. Due to the high number of ruck duels in a game, ruckmen recorded the highest number of game activities (173) but, if these are subtracted, midfielders and ruckmen on average each had more than 100 (119 and 106 respectively) game activities, with CHF/CHB averaging 92 and SF/B and FF/FB 75 and 71 each. As a frequency, ruckmen have a game involvement every 45 secs, midfielders every 63 secs, CHF/CHB every 83 secs, SF/B every 95 secs and FF/FB every 112 secs.

285


Position

Category

Marks Taken Marl

Overhead Chest Overhead Chest Mark Overhead Mark Chest Mark Overhead Chest Overh Mark Mark Mark Contested Contested Uncontested Uncontested Mark Mark Mark Co,

FF/FB 2 2 1 2 2 0 11 1 1C (1-4) (1-5) (0-1) (0-5) (0-3) (0-1) (6-14) (0-2) (6-1

CHF/CHB 3 3 1 0 2 2 14 3 13 (1-4) (1-4) (0-1) (0-1) (1-3) (1-4) (8-23)( 2-3) (8-2

Ruck 1 2 1 1 1 1 12 1 12 (0-2) (0-5) (0-1) (0-3) (0-1) (0-2) (7-15) (0-1) (7-1

Small F/B 2 4 1 1 1 3 4 1 4 (2-2) (1-7) (0-2) (0-2) (0-2) (1-6) (2-6) (0-1) (2-{

Midfield 2 2 1 1 1 1 3 0 3 (0-3) (0-3) (0-2) (0-1) (0-3) (0-2) (1-6) (0-1) (1-{

FF = full forward, FB = full back, F = forward, B = back, CHF = centre half forward, CriB = centre half back

Table 4: Game activities for different positions (marks, kick and handballs). Means and ranges (in brackets) are~nted.

Position Ruck Duels

Category Bounce Downs ThrOw InS Bounce Downs Throw Ins Tackles TaCkleS Shepherds Shepherds Sm contested Contested Uncontested Uncontested Given Received Given Received

FF/FB 0 2 0 0 1 2 1 1 (0-0) (0-5) (0-0) (0-0) (0-3) (1-3) (1-2) (0-2) (

CHF/CHB 1 5 0 0 5 2 2 1 (0-4) (1-4) (0-0) (0-0) (1-12) (0-4) (2-3) (0-1) (

Ruck 49 18 0 0 8 5 2 2 (41-57) (12-29) (0-0) (0-0) (5-10) (3-7) 0-4) (1-4) (

Small F/B 0 0 0 0 4 5 1 1 (0-0) (0-1) (0-0) (0-0) (1-6) (3-6) (1-2) (0-I) (

Midfield 0 1 0 0 10 8 2 0 (0-1) (0-2) (0-0) (0-0) (5-15) (6-12) (0-4) (0-0) ('

FF = full forward, FB = full back, F = forward, B = back, CHF = centre half forward, Crib = centre half back

Table 5: Game activities for different positions (Ruck duels and team involvements). Means and ranges (in brac~re presen

Position Total Number of Frequency of Game Category Game Activities Activities (per second)

FF/FB 71 _+ 19 1 per 112 (52-91) (79-149)

Midfield 119 + 17 1 per 63 (100-150) (51-75)

Ruck 173 _+ 36 1 per 45 (148-215) (36-48)

Small F/B 75 _+ 11 1 per 95 (60-84) (86-109)

CHF/CHB 92 +_ 20 1 per 83 (67-115) (63-111)

FF = full forward, FB = full back, F = forward, B = back, CHF = centre half forward, CHB = centre half back.

Table 6: Total game activities for different positions, including a calculation of the frequency of occurrence during a match. Means ( +-SD) and ranges (in brackets) are presented.

286


Marks MISSed Kicks Handballs

Overhead Chest Mark Overhead Mark Chest Mark Total KiCkS KICKS TOtal Handballs Handballs Mark Contested Contested Uncontested Uncontested Kicks Contested Uncontested Handballs Contested uncontested

10 1 0 0 9 2 7 4 1 3 (6-13) (0-2) (0-1) (0-1) (6-12) (0-3) (3-11) (0-9) (0-2) (0-7)

13 2 0 0 8 1 7 7 5 3 (8-22) (2-3) (0-1) (0-1) (7-10) (1-2) (5-9) (4-11) (2-7) (0-4)

12 0 0 0 11 4 6 7 5 2 (7-15) (0-1) (0-1) (0-1) (6-13) (1-9) (3-12) (5-10) (2-10) (0-3)

4 1 0 0 13 2 11 6 3 3 (2-6) (0-1) (0-0) (0-0) (8-15) (1-3) (7-14) (1-11) (0-8) (0-6)

3 0 0 0 13 5 8 9 6 3 (1-6) (0-1) (0-0) (0-0) (10-16) (3-6) (5-11) (4-12) (3-7) (1-8)

rated.

Team Involvements

Smothers Smothers Spoils Spoils Bumps Bumps Knock Ons Going to Ground Given Received Given Received Given Received Given (No. of times)

3 0 4 2 3 2 1 8 (1-6) (0-0) (1-10) (0-4) (2-4) (1-3) (0-3) (5-13)

4 0 6 2 2 4 1 11 (0-7) (0-1) (1-11) (0-7) (1-4) (2-4) (0-2) (8-17)

4 1 5 1 4 3 1 23 (2-6) (1-1) (2-10) (0-2) (2-6) (3-3) (0-1) (15-34)

2 1 1 1 2 3 1 11 (1-3) (0-1) (0-3) (0-2) (0-6) (0-7) (0-2) (8-15)

5 1 2 1 3 6 1 21 (1-10) (0-2] (1-3) (0-2) (0-4) (2-10) (0-5) (7-28)

c~re presented. Position Total <lsec l-2sec 2-4 sec 4-Gsec >6sec

FF/FB 8 _+ 2 2 3 2 1 0 (3-15) (0-3) (1-7) (0-4) (0-2) (0-0)

Midfield 21 _+ 9 7 5 7 2 0 (9-31) (2-18) (2-8) (3-11) (0-8) (0-0)

Ruck 6 _+ 3 1 3 2 0 0 (4-11) (0-2) (1-5) (I -3) (0-1) (0-1)

Small F/B 11 _+ 3 3 3 4 1 1 (8-15) (1-3) (1-4) (3-6) (0-2) (0-2)

CHF/CHB 10 _+ 5 3 3 3 1 0 (4-16) (2 -6) (0-4) (1-7) (0-2) (0-0)

FF = full forward, FB = full back, F = forward, B = back, CHF = centre half forward, CHB = centre half back.

Table 7: Number and ranges of time recorded for ground ball contests in games. Means ( +_SD) and ranges (in brackets) are presented.

287

Player movement patterns and game activities,..

Table 7 repor ts on the n u m b e r and range of t imes recorded for ground ball contes ts in games. Midfielders recorded, on average, 2-3 t imes as m a n y ground ball contes ts as for every other position (21 versus 7-11). Regardless of position, however, virtually all the ground ball contes ts took less t han 6 secs, with approximate ly 90% being 4 secs or less in durat ion.

DiSCUSSiOn Norton et al 4 concluded their recent paper on the evolution of Austra l ian Football by stat ing that: "We now need detailed video analyses on individual players during games to complete the picture of game demands". The presen t s tudy has achieved this by filming the same individual players (from different positional categories) in two complete games each in the 2000 AFL season.

Movement Patterns Of the five positional types investigated, FF/FB appeared mos t different from the other positions, as they were seen to s tand more and jog and fas t - run less. The players f l m e d in these posit ions were acknowledged "specialists" in these roles, as they were rarely ever moved to another position within a game. If a club has "specialist" FF/FB players on its list, they probably war ran t more specific position t raining t han other players. For high intensi ty conditioning work, relatively more spr int movements , with relatively less fast-running, would seem appropriate , which should be combined with game activities specific to the position (eg, leading, contested marking, etc). Of the other positions, midfielders were found to do more fas t - running t han other players, which may underscore their greater a m o u n t of game activities and ground ball contests , as they cover more ground at speed in order to get to more ball contests. For specific game fitness, midfielders might be given additional fast- running beyond tha t prescr ibed for other players. However, all players should do repeated short, fas t - running (and sprinting) efforts to build game fitness. Mixed speed runs , in which fas t - running precedes a n d / o r follows sprint ing m a y be an effective training method, as in the games analysed mos t spr ints were preceded by a period of fast-running, before accelerating to maximal speeds.

Midfielders were found to cover the greates t es t imated total dis tance in games, being slightly less t han 17,000 m, with the FF/FB being the least, at j u s t over 13,600 m. Ruckmen covered nearly 15,400 m and the CHF/CHB and SF/B bo th were es t imated to cover j u s t over 16,000 m in games. These es t imated dis tances concur with the resul ts found by Norton et ale us ing real t ime compute r t racking of players in AFL matches (eg, 20 midfielders were recorded as covering a lmost 17,500 m on average). The total d is tances covered in the modern game are considerably greater than previously reported (visual recordings) by Pyke and Smith 9 (for one half; cen t r eman - 6.6 km, half back flanker - 5 kin), Hahn et al2 (rovers, 10-11 km) and Douge 1° (forwards 4000 m, ruck-rovers 6800 m). There is little doubt tha t cur ren t players require superior aerobic fitness in order to cover such large dis tances during a game, and this requi rement is not confined to the more nomadic midfielders, as the posit ion players (CHF/CHB and SF/B) were also seen to cover more t h a n 16,000 m in games. More recent tactical innovations in the game, such as set rotat ion of players (especially midfielders) on and off the ground and flooding the

288


opposition forward line may also change the total dis tances covered by players during games. Playing in more t han one position during a game will also have an influence on this variable.

Although the total dis tances covered in games by players are popularised by the media, the a m o u n t and durat ion of high intensi ty running is of more practical impor tance and significance to game performance. High intensity (fast-running and sprinting) efforts accounted for only 4-6% of the total movement t ime in games bu t there are more than 150 efforts of this sort in a game for all positions. Previously, Hahn et al2 and McKenna et al 3 reported an average of 110-115 and 98 high intensi ty runs in a ma tch for rovers. Fur thermore , virtually all of the high intensity efforts lasted for less t han 6 secs, so conditioning drills should be s t ruc tured accordingly. This brief duration is consis tent with previous fnd ings for Austral ian Football 1,3,s, suggesting tha t the durat ion of high intensi ty running has not changed from the 1970s to the present day. Repeated fas t - running and sprinting efforts of 6 secs or less should be used to build specific game fitness, in terspersed with walking and jogging, as players spend the majori ty of their recovery t ime between high intensi ty efforts in these movement pat terns . The recovery periods between high intensity (fast-running or sprinting) movements in games are relatively shor t (40-50 secs) for all posit ions except the FF/FB (70 secs). These figures are somewhat lower than reported previously for Austral ian Football rovers (73 secs) 3. However, mean values can be potentially misleading. Of more relevance, nearly half are 0-20 secs in duration, and approximately 70% are 0-60 secs in length. Therefore, in conditioning drills (for all positions), recovery periods between high intensity movements and game activities such as ground ball contests, (which generally also only last for up to 6 secs), should most ly be brief (<20 secs) and not more than 60 secs, in order to ma tch game demands and provide overload.

Of fur ther interest is that , of the spr ints recorded here in games, more than half involved at least one change of direction, which was most ly in the arc of 0- 90 ° (left or right). Changes of direction are frequent in the game of Austral ian Football, bu t have not previously been documented. It is appa ren t tha t spr int training should not be conducted in s t raight lines only. Changes of direction, incorporat ing left and right lateral movements (mostly 0-90 ° ) should be regularly p rogrammed into spr int t raining drills for Austral ian footballers.

Game Activities Of the five posit ions studied, r uckm en and midfielders have the mos t game activities and the higher frequencies of involvement ( ruckmen are 1 per 45 secs, midfielders are 1 per 63 secs), and these position types should have specific training drills designed for them, which involve several repeats of common game activities (eg, for ruckmen; ruck duels, contested marking, going to ground, etc and, for midfielders, ground ball contests, tackling, going to ground, etc) i n t e r spe r sed with jogging, f a s t - r u n n i n g a n d / o r spr int ing movements to move to different par ts of the ground, before performing the next game activity, and so on. The n u m b e r of game activities recorded for midfielders (119) and r u c k m e n (106) were somewhat higher t han the n u m b e r (85) reported previously for Austral ian Football rovers 3. As a result, the t ime between high intensity efforts found here for midfielders (41 secs) is m u c h lower than their

289


finding for rovers (73 secs). When planning training sess ions for r u c k m e n and midfielders, it is impor tan t to consider tha t the clearance ra tes from ruck contes ts have not been shown to be associated with winning the game or quar te r 5. However, it would still provide specific player training for improving the t eam ' s chances of winning initial possess ion of the ball after bounce-downs and throw-ins. A similar style of drill could be used for the other positional types also, as training drills tha t combine specific game activities with spec i fc movemen t pa t t e rns should be encouraged. With regard to ruck duels, the da ta reported in Table 5 do not represent all of the ruck contes ts within a game, as tall key position players usual ly contest abou t 8% of bounce-downs and 26% of throw-ins in a game 5, and this s i tuat ion should also be considered for t raining purposes .

For the taller players, m a n y more m a r k s are missed t han taken. This is consis tent with recent AFL game analysis 5, which reported a lmost half of all mark ing contests (opposed or unopposed) are unsuccess fu l and tha t spoiling dur ing opposed contes ts only accounted for abou t 20% of the unsuccess fu l a t tempts . While the resul t found here m a y reflect good defensive skills (ie, effective spoiling by an opponent), it m a y also reflect tha t contested overhead m a r k s may not be pract ised sufficiently a t t raining 5.

In simple terms, Austra l ian Football can be described as a series of contests for the ball between small groups of opposing players and m a n y of these involve ground balls, which for this s tudy was defined as any contes t for the ball other t han a ruck duel or mark ing contest . Here, ground ball contes ts were most ly less t han 4 secs in durat ion, and all took less t han 6 secs. Training drills involving ground ball contests should have players per form several (5-10) shor t (<6-sec) efforts, in te rspersed with some walk / jog recovery (15-30 secs, depending on n u m b e r and t ime of efforts) to provide overload, ra ther t han one long (30-60 sec) cont inuous effort, as this is not specific to game demands . Training activities should include a combinat ion of ground ball and mark ing contests. However, midfield p l a y e r s have 2-3 t imes as m a n y ground ball contests as the other positions, so specific drills for them should be more frequently p rogrammed than for other players.

Lastly, a common occurrence in contact games, which has not previously been reported in Austra l ian Football, is "going to ground". Players went to ground on average 8-23 t imes per game, with rucks and midfielders having the greates t tallies. Every t ime players go to ground they m u s t expend energy in g e t t i n g u p off the ground in order to run again, so specific game conditioning drills should incorporate this common game activity.

In conclusion, this video analysis of Austral ian Football ha s provided an improved unders tand ing of the game demands for players, which should enable coaches to bet ter t rain their players to play the game. Improvements in p re - season and in-season training pract ices should be possible after careful interpretat ion of the resul ts of this study, including more specific training rout ines for certain positional roles and individual players. In brief, t raining for Austra l ian Football players should closely replicate the in termit tent na ture of the game. High intensity efforts accoun t for only 4-6% of total movemen t t ime in games, bu t there are more t h a n 150 of these efforts during a game, virtually all last ing for less t h a n 6 secs. Consequently, t raining should incorporate a

290


rapid turnover of both high intensity efforts and game-related activities, mostly of a very short duration, interspersed with longer periods of low intensity effort for recovery. Repeated fast running and sprinting efforts of 6 secs or less interspersed with walking and jogging movements should be used to build specific game fitness.

Acknowledgements The generous assistance of the AFL in funding this project is gratefully acknowledged, as is the cooperation of the West Coast Eagles and Fremantle Football clubs. In addition, the computing expertise of Ken Godwin (West Coast Eagles) who gave up considerable time for software design, set-up and troubleshooting is greatly appreciated.

References 1. Jaques TD and Pavia GR. An analysis of the movement pat terns of players in an Australian Rules league football match. Aust J Sport Med 1974;5:10-21. 2. Hahn A, Taylor N, Hunt B, Woodhouse T. and Shultz G. Physiological relationships between training activities and match play in Australian Football rovers. Sports Coach 1979;3(3):3-8. 3. McKenna MJ, Patrick JD, Sandstrom ER and Chennells MHD. Computer-video analysis of activity pat terns in Australian Rules football. In Reilly et al (eds), Science and Football: proceedings of the First World Congress o f Science and Football (274-281). Liverpool, E. & FN Spon. 1988. 4. Norton KI, Craig NP and Olds TS. The evolution of Australian Football. J Sci Med Sport 1999;2(4):389-404. 5. Appleby B. and Dawson B. Analysis of selected game activities in Australian Football. J Sci Med Sport 2002;5(2): 129-142. 6. Mayhew JL and Wenger HA. Time motion analysis of professional soccer. J Hum Mvt Studies 1985; 11:49-52. 7. Durward, B.R., Baer, G.D. and Rowe, P.J. Functional Human Movement: Measurement and Analysis. Melbourne, Butterworth Heinemann. 1996. 8. Norton K Schwerdt S. and Craig N. Player movement and game structure in the Australian Football League. Australian Football League Research Report. 2001. 9. Pyke F. and Smith R. Football: The Scientific Way. University of Western Australia Press. 1975. 10. Douge B. Testing in Australian Rules football. Sports Coach. 1982;6:29-37.

291

player movement patterns and game activities in the australian football league

Documents