correlation through bounded memory strategies...correlation through bounded memory strategies ron...

76
. . . . . . . . Correlation Through Bounded Memory Strategies Ron Peretz 1 (joint work with Olivier Gossner and Pen´ elope Hern´ andez) 1 Tel Aviv University November, 2011 Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 1 / 14

Upload: others

Post on 11-Mar-2020

18 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

.

......Correlation Through Bounded Memory Strategies

Ron Peretz1

(joint work with Olivier Gossner and Penelope Hernandez)

1Tel Aviv University

November, 2011

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 1 / 14

Page 2: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Introductory example

Take a random ordering of the integers 0, 1, . . . , 7.

Write them in binary representation. We have a random sequence of24 bits, x1, . . . , x24.

How random is this sequence? How well does it play repeatedmatching pennies?

The entropy method shows that if we take n2n such bits theguaranteed value converges to zero.

What is the value if we are allowed to take a glimpse at therealization of x1, . . . , xn2n?

Bounded memory.

.Let’s play!........

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 2 / 14

Page 3: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Introductory example

Take a random ordering of the integers 0, 1, . . . , 7.

Write them in binary representation. We have a random sequence of24 bits, x1, . . . , x24.

How random is this sequence? How well does it play repeatedmatching pennies?

The entropy method shows that if we take n2n such bits theguaranteed value converges to zero.

What is the value if we are allowed to take a glimpse at therealization of x1, . . . , xn2n?

Bounded memory.

.Let’s play!........

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 2 / 14

Page 4: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Introductory example

Take a random ordering of the integers 0, 1, . . . , 7.

Write them in binary representation. We have a random sequence of24 bits, x1, . . . , x24.

How random is this sequence? How well does it play repeatedmatching pennies?

The entropy method shows that if we take n2n such bits theguaranteed value converges to zero.

What is the value if we are allowed to take a glimpse at therealization of x1, . . . , xn2n?

Bounded memory.

.Let’s play!........

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 2 / 14

Page 5: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Introductory example

Take a random ordering of the integers 0, 1, . . . , 7.

Write them in binary representation. We have a random sequence of24 bits, x1, . . . , x24.

How random is this sequence? How well does it play repeatedmatching pennies?

The entropy method shows that if we take n2n such bits theguaranteed value converges to zero.

What is the value if we are allowed to take a glimpse at therealization of x1, . . . , xn2n?

Bounded memory.

.Let’s play!........

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 2 / 14

Page 6: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Introductory example

Take a random ordering of the integers 0, 1, . . . , 7.

Write them in binary representation. We have a random sequence of24 bits, x1, . . . , x24.

How random is this sequence? How well does it play repeatedmatching pennies?

The entropy method shows that if we take n2n such bits theguaranteed value converges to zero.

What is the value if we are allowed to take a glimpse at therealization of x1, . . . , xn2n?

Bounded memory.

.Let’s play!........1

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 2 / 14

Page 7: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Introductory example

Take a random ordering of the integers 0, 1, . . . , 7.

Write them in binary representation. We have a random sequence of24 bits, x1, . . . , x24.

How random is this sequence? How well does it play repeatedmatching pennies?

The entropy method shows that if we take n2n such bits theguaranteed value converges to zero.

What is the value if we are allowed to take a glimpse at therealization of x1, . . . , xn2n?

Bounded memory.

.Let’s play!........10

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 2 / 14

Page 8: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Introductory example

Take a random ordering of the integers 0, 1, . . . , 7.

Write them in binary representation. We have a random sequence of24 bits, x1, . . . , x24.

How random is this sequence? How well does it play repeatedmatching pennies?

The entropy method shows that if we take n2n such bits theguaranteed value converges to zero.

What is the value if we are allowed to take a glimpse at therealization of x1, . . . , xn2n?

Bounded memory.

.Let’s play!........101

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 2 / 14

Page 9: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Introductory example

Take a random ordering of the integers 0, 1, . . . , 7.

Write them in binary representation. We have a random sequence of24 bits, x1, . . . , x24.

How random is this sequence? How well does it play repeatedmatching pennies?

The entropy method shows that if we take n2n such bits theguaranteed value converges to zero.

What is the value if we are allowed to take a glimpse at therealization of x1, . . . , xn2n?

Bounded memory.

.Let’s play!........101 0

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 2 / 14

Page 10: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Introductory example

Take a random ordering of the integers 0, 1, . . . , 7.

Write them in binary representation. We have a random sequence of24 bits, x1, . . . , x24.

How random is this sequence? How well does it play repeatedmatching pennies?

The entropy method shows that if we take n2n such bits theguaranteed value converges to zero.

What is the value if we are allowed to take a glimpse at therealization of x1, . . . , xn2n?

Bounded memory.

.Let’s play!........101 00

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 2 / 14

Page 11: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Introductory example

Take a random ordering of the integers 0, 1, . . . , 7.

Write them in binary representation. We have a random sequence of24 bits, x1, . . . , x24.

How random is this sequence? How well does it play repeatedmatching pennies?

The entropy method shows that if we take n2n such bits theguaranteed value converges to zero.

What is the value if we are allowed to take a glimpse at therealization of x1, . . . , xn2n?

Bounded memory.

.Let’s play!........101 001

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 2 / 14

Page 12: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Introductory example

Take a random ordering of the integers 0, 1, . . . , 7.

Write them in binary representation. We have a random sequence of24 bits, x1, . . . , x24.

How random is this sequence? How well does it play repeatedmatching pennies?

The entropy method shows that if we take n2n such bits theguaranteed value converges to zero.

What is the value if we are allowed to take a glimpse at therealization of x1, . . . , xn2n?

Bounded memory.

.Let’s play!........101 001 111 011 000 100 110

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 2 / 14

Page 13: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Introductory example

Take a random ordering of the integers 0, 1, . . . , 7.

Write them in binary representation. We have a random sequence of24 bits, x1, . . . , x24.

How random is this sequence? How well does it play repeatedmatching pennies?

The entropy method shows that if we take n2n such bits theguaranteed value converges to zero.

What is the value if we are allowed to take a glimpse at therealization of x1, . . . , xn2n?

Bounded memory.

.Let’s play!........101 001 111 011 000 100 110 010

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 2 / 14

Page 14: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Introductory example

Take a random ordering of the integers 0, 1, . . . , 7.

Write them in binary representation. We have a random sequence of24 bits, x1, . . . , x24.

How random is this sequence? How well does it play repeatedmatching pennies?

The entropy method shows that if we take n2n such bits theguaranteed value converges to zero.

What is the value if we are allowed to take a glimpse at therealization of x1, . . . , xn2n?

Bounded memory.

.Let’s play!........101 001 111 011 000 100 110 010

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 2 / 14

Page 15: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Introductory example

Take a random ordering of the integers 0, 1, . . . , 7.

Write them in binary representation. We have a random sequence of24 bits, x1, . . . , x24.

How random is this sequence? How well does it play repeatedmatching pennies?

The entropy method shows that if we take n2n such bits theguaranteed value converges to zero.

What is the value if we are allowed to take a glimpse at therealization of x1, . . . , xn2n?

Bounded memory.

.

Let’s play!

..

......

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 2 / 14

Page 16: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Introductory example

Take a random ordering of the integers 0, 1, . . . , 7.

Write them in binary representation. We have a random sequence of24 bits, x1, . . . , x24.

How random is this sequence? How well does it play repeatedmatching pennies?

The entropy method shows that if we take n2n such bits theguaranteed value converges to zero.

What is the value if we are allowed to take a glimpse at therealization of x1, . . . , xn2n?

Bounded memory.

.Let’s play again!........

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 2 / 14

Page 17: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Introductory example

Take a random ordering of the integers 0, 1, . . . , 7.

Write them in binary representation. We have a random sequence of24 bits, x1, . . . , x24.

How random is this sequence? How well does it play repeatedmatching pennies?

The entropy method shows that if we take n2n such bits theguaranteed value converges to zero.

What is the value if we are allowed to take a glimpse at therealization of x1, . . . , xn2n?

Bounded memory.

.Let’s play again!........1

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 2 / 14

Page 18: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Introductory example

Take a random ordering of the integers 0, 1, . . . , 7.

Write them in binary representation. We have a random sequence of24 bits, x1, . . . , x24.

How random is this sequence? How well does it play repeatedmatching pennies?

The entropy method shows that if we take n2n such bits theguaranteed value converges to zero.

What is the value if we are allowed to take a glimpse at therealization of x1, . . . , xn2n?

Bounded memory.

.Let’s play again!........10

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 2 / 14

Page 19: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Introductory example

Take a random ordering of the integers 0, 1, . . . , 7.

Write them in binary representation. We have a random sequence of24 bits, x1, . . . , x24.

How random is this sequence? How well does it play repeatedmatching pennies?

The entropy method shows that if we take n2n such bits theguaranteed value converges to zero.

What is the value if we are allowed to take a glimpse at therealization of x1, . . . , xn2n?

Bounded memory.

.Let’s play again!........101

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 2 / 14

Page 20: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Introductory example

Take a random ordering of the integers 0, 1, . . . , 7.

Write them in binary representation. We have a random sequence of24 bits, x1, . . . , x24.

How random is this sequence? How well does it play repeatedmatching pennies?

The entropy method shows that if we take n2n such bits theguaranteed value converges to zero.

What is the value if we are allowed to take a glimpse at therealization of x1, . . . , xn2n?

Bounded memory.

.Let’s play again!........101 0

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 2 / 14

Page 21: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Introductory example

Take a random ordering of the integers 0, 1, . . . , 7.

Write them in binary representation. We have a random sequence of24 bits, x1, . . . , x24.

How random is this sequence? How well does it play repeatedmatching pennies?

The entropy method shows that if we take n2n such bits theguaranteed value converges to zero.

What is the value if we are allowed to take a glimpse at therealization of x1, . . . , xn2n?

Bounded memory.

.Let’s play again!........101 00

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 2 / 14

Page 22: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Introductory example

Take a random ordering of the integers 0, 1, . . . , 7.

Write them in binary representation. We have a random sequence of24 bits, x1, . . . , x24.

How random is this sequence? How well does it play repeatedmatching pennies?

The entropy method shows that if we take n2n such bits theguaranteed value converges to zero.

What is the value if we are allowed to take a glimpse at therealization of x1, . . . , xn2n?

Bounded memory.

.Let’s play again!........101 001

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 2 / 14

Page 23: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Introductory example

Take a random ordering of the integers 0, 1, . . . , 7.

Write them in binary representation. We have a random sequence of24 bits, x1, . . . , x24.

How random is this sequence? How well does it play repeatedmatching pennies?

The entropy method shows that if we take n2n such bits theguaranteed value converges to zero.

What is the value if we are allowed to take a glimpse at therealization of x1, . . . , xn2n?

Bounded memory.

.Let’s play again!........101 001 1

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 2 / 14

Page 24: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Introductory example

Take a random ordering of the integers 0, 1, . . . , 7.

Write them in binary representation. We have a random sequence of24 bits, x1, . . . , x24.

How random is this sequence? How well does it play repeatedmatching pennies?

The entropy method shows that if we take n2n such bits theguaranteed value converges to zero.

What is the value if we are allowed to take a glimpse at therealization of x1, . . . , xn2n?

Bounded memory.

.Let’s play again!........101 001 11

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 2 / 14

Page 25: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Introductory example

Take a random ordering of the integers 0, 1, . . . , 7.

Write them in binary representation. We have a random sequence of24 bits, x1, . . . , x24.

How random is this sequence? How well does it play repeatedmatching pennies?

The entropy method shows that if we take n2n such bits theguaranteed value converges to zero.

What is the value if we are allowed to take a glimpse at therealization of x1, . . . , xn2n?

Bounded memory.

.Let’s play again!........101 001 111

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 2 / 14

Page 26: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Introductory example

Take a random ordering of the integers 0, 1, . . . , 7.

Write them in binary representation. We have a random sequence of24 bits, x1, . . . , x24.

How random is this sequence? How well does it play repeatedmatching pennies?

The entropy method shows that if we take n2n such bits theguaranteed value converges to zero.

What is the value if we are allowed to take a glimpse at therealization of x1, . . . , xn2n?

Bounded memory.

.Let’s play again!........101 001 111 0

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 2 / 14

Page 27: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Introductory example

Take a random ordering of the integers 0, 1, . . . , 7.

Write them in binary representation. We have a random sequence of24 bits, x1, . . . , x24.

How random is this sequence? How well does it play repeatedmatching pennies?

The entropy method shows that if we take n2n such bits theguaranteed value converges to zero.

What is the value if we are allowed to take a glimpse at therealization of x1, . . . , xn2n?

Bounded memory.

.Let’s play again!........101 001 111 01...

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 2 / 14

Page 28: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Introductory example

Take a random ordering of the integers 0, 1, . . . , 7.

Write them in binary representation. We have a random sequence of24 bits, x1, . . . , x24.

How random is this sequence? How well does it play repeatedmatching pennies?

The entropy method shows that if we take n2n such bits theguaranteed value converges to zero.

What is the value if we are allowed to take a glimpse at therealization of x1, . . . , xn2n?

Bounded memory.

.Let’s play again!........101 001 111 01...

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 2 / 14

Page 29: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

The model – bounded memory

We consider the (undiscounted) repeated version of a finite game instrategic form G = ⟨A =

∏i∈N Ai , g : A → RN⟩.

An m-memory strategy τ is one that satisfies

#{τ|h : “h is a finite history”

}≤ m,

where τ|h is the strategy that τ induces on the sub-game originatingat h.

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 3 / 14

Page 30: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

The model – bounded memory

We consider the (undiscounted) repeated version of a finite game instrategic form G = ⟨A =

∏i∈N Ai , g : A → RN⟩.

An m-memory strategy τ is one that satisfies

#{τ|h : “h is a finite history”

}≤ m,

where τ|h is the strategy that τ induces on the sub-game originatingat h.

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 3 / 14

Page 31: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Motivation

Consider the (undiscounted infinitely) repeated game G (m1, . . . ,mN)in which each player i is restricted to mi -memory strategies.

We would like to be able to say intelligible things about its “solution.”

For example,

- min-max and max-min (one maximizer against N − 1 minimizers),

- mixed and pure,

- individually rational level (mixed min-max) and equilibrium payoffs(folk theorem).

Those are tough problems.

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 4 / 14

Page 32: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Motivation

Consider the (undiscounted infinitely) repeated game G (m1, . . . ,mN)in which each player i is restricted to mi -memory strategies.

We would like to be able to say intelligible things about its “solution.”

For example,

- min-max and max-min (one maximizer against N − 1 minimizers),

- mixed and pure,

- individually rational level (mixed min-max) and equilibrium payoffs(folk theorem).

Those are tough problems.

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 4 / 14

Page 33: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Motivation

Consider the (undiscounted infinitely) repeated game G (m1, . . . ,mN)in which each player i is restricted to mi -memory strategies.

We would like to be able to say intelligible things about its “solution.”

For example,

- min-max and max-min (one maximizer against N − 1 minimizers),

- mixed and pure,

- individually rational level (mixed min-max) and equilibrium payoffs(folk theorem).

Those are tough problems.

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 4 / 14

Page 34: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Motivation

Consider the (undiscounted infinitely) repeated game G (m1, . . . ,mN)in which each player i is restricted to mi -memory strategies.

We would like to be able to say intelligible things about its “solution.”

For example,

- min-max and max-min (one maximizer against N − 1 minimizers),

- mixed and pure,

- individually rational level (mixed min-max) and equilibrium payoffs(folk theorem).

Those are tough problems.

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 4 / 14

Page 35: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Motivation

Consider the (undiscounted infinitely) repeated game G (m1, . . . ,mN)in which each player i is restricted to mi -memory strategies.

We would like to be able to say intelligible things about its “solution.”

For example,

- min-max and max-min (one maximizer against N − 1 minimizers),

- mixed and pure,

- individually rational level (mixed min-max) and equilibrium payoffs(folk theorem).

Those are tough problems.

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 4 / 14

Page 36: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Motivation

Consider the (undiscounted infinitely) repeated game G (m1, . . . ,mN)in which each player i is restricted to mi -memory strategies.

We would like to be able to say intelligible things about its “solution.”

For example,

- min-max and max-min (one maximizer against N − 1 minimizers),

- mixed and pure,

- individually rational level (mixed min-max) and equilibrium payoffs(folk theorem).

Those are tough problems.

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 4 / 14

Page 37: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Motivation

Consider the (undiscounted infinitely) repeated game G (m1, . . . ,mN)in which each player i is restricted to mi -memory strategies.

We would like to be able to say intelligible things about its “solution.”

For example,

- min-max and max-min (one maximizer against N − 1 minimizers),

- mixed and pure,

- individually rational level (mixed min-max) and equilibrium payoffs(folk theorem).

Those are tough problems.

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 4 / 14

Page 38: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Matching Pennies

1 0

0 1valG = 1

2puremaxminG = 0

Ben Porath (1988).lim inf valG (m, 2o(m)) ≥ 1

2 .

Neyman (1998). If limG (m, 2m/o(1)) = 0.

Neyman-Spencer (2007).lim valG (m, 2(1−o(1))m) = 0.

Neyman (2008).lim inf G (m, 2Θm) ≥ H−1(1−Θ).

limm→∞ valG (m, 2Θm) exists?

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 5 / 14

Page 39: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Matching Pennies

1 0

0 1valG = 1

2puremaxminG = 0

Ben Porath (1988).lim inf valG (m, 2o(m)) ≥ 1

2 .

Neyman (1998). If limG (m, 2m/o(1)) = 0.

Neyman-Spencer (2007).lim valG (m, 2(1−o(1))m) = 0.

Neyman (2008).lim inf G (m, 2Θm) ≥ H−1(1−Θ).

limm→∞ valG (m, 2Θm) exists?

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 5 / 14

Page 40: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Matching Pennies

1 0

0 1valG = 1

2puremaxminG = 0

Ben Porath (1988).lim inf valG (m, 2o(m)) ≥ 1

2 .

Neyman (1998). If limG (m, 2m/o(1)) = 0.

Neyman-Spencer (2007).lim valG (m, 2(1−o(1))m) = 0.

Neyman (2008).lim inf G (m, 2Θm) ≥ H−1(1−Θ).

limm→∞ valG (m, 2Θm) exists?

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 5 / 14

Page 41: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Matching Pennies

1 0

0 1valG = 1

2puremaxminG = 0

Ben Porath (1988).lim inf valG (m, 2o(m)) ≥ 1

2 .

Neyman (1998). If limG (m, 2m/o(1)) = 0.

Neyman-Spencer (2007).lim valG (m, 2(1−o(1))m) = 0.

Neyman (2008).lim inf G (m, 2Θm) ≥ H−1(1−Θ).

limm→∞ valG (m, 2Θm) exists?

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 5 / 14

Page 42: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Matching Pennies

1 0

0 1valG = 1

2puremaxminG = 0

Ben Porath (1988).lim inf valG (m, 2o(m)) ≥ 1

2 .

Neyman (1998). If limG (m, 2m/o(1)) = 0.

Neyman-Spencer (2007).lim valG (m, 2(1−o(1))m) = 0.

Neyman (2008).lim inf G (m, 2Θm) ≥ H−1(1−Θ).

limm→∞ valG (m, 2Θm) exists?

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 5 / 14

Page 43: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Matching Pennies

1 0

0 1valG = 1

2puremaxminG = 0

Ben Porath (1988).lim inf valG (m, 2o(m)) ≥ 1

2 .

Neyman (1998). If limG (m, 2m/o(1)) = 0.

Neyman-Spencer (2007).lim valG (m, 2(1−o(1))m) = 0.

Neyman (2008).lim inf G (m, 2Θm) ≥ H−1(1−Θ).

limm→∞ valG (m, 2Θm) exists?

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 5 / 14

Page 44: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Self advertisement

For N = 2, there are solutions to restricted models of boundedmemory.

- Neyman (2008) and Peretz (forthcoming MOR) solveG (moblivious , 2

Θm),

- Peretz (2011 GEB) solves G (krecall ,mrecall).

For N = 3,

- Peretz (forthcoming IJGT) suggests some bounds for the mixedmin-max of G (Θmrecall ,mrecall ,mrecall).

- The present work facilitates bounds for the mixed max-min ofG (2Θm logm,m,m) and pure min-max of G (Θm,m,m).

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 6 / 14

Page 45: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Self advertisement

For N = 2, there are solutions to restricted models of boundedmemory.

- Neyman (2008) and Peretz (forthcoming MOR) solveG (moblivious , 2

Θm),

- Peretz (2011 GEB) solves G (krecall ,mrecall).

For N = 3,

- Peretz (forthcoming IJGT) suggests some bounds for the mixedmin-max of G (Θmrecall ,mrecall ,mrecall).

- The present work facilitates bounds for the mixed max-min ofG (2Θm logm,m,m) and pure min-max of G (Θm,m,m).

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 6 / 14

Page 46: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Self advertisement

For N = 2, there are solutions to restricted models of boundedmemory.

- Neyman (2008) and Peretz (forthcoming MOR) solveG (moblivious , 2

Θm),

- Peretz (2011 GEB) solves G (krecall ,mrecall).

For N = 3,

- Peretz (forthcoming IJGT) suggests some bounds for the mixedmin-max of G (Θmrecall ,mrecall ,mrecall).

- The present work facilitates bounds for the mixed max-min ofG (2Θm logm,m,m) and pure min-max of G (Θm,m,m).

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 6 / 14

Page 47: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Self advertisement

For N = 2, there are solutions to restricted models of boundedmemory.

- Neyman (2008) and Peretz (forthcoming MOR) solveG (moblivious , 2

Θm),

- Peretz (2011 GEB) solves G (krecall ,mrecall).

For N = 3,

- Peretz (forthcoming IJGT) suggests some bounds for the mixedmin-max of G (Θmrecall ,mrecall ,mrecall).

- The present work facilitates bounds for the mixed max-min ofG (2Θm logm,m,m) and pure min-max of G (Θm,m,m).

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 6 / 14

Page 48: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Self advertisement

For N = 2, there are solutions to restricted models of boundedmemory.

- Neyman (2008) and Peretz (forthcoming MOR) solveG (moblivious , 2

Θm),

- Peretz (2011 GEB) solves G (krecall ,mrecall).

For N = 3,

- Peretz (forthcoming IJGT) suggests some bounds for the mixedmin-max of G (Θmrecall ,mrecall ,mrecall).

- The present work facilitates bounds for the mixed max-min ofG (2Θm logm,m,m) and pure min-max of G (Θm,m,m).

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 6 / 14

Page 49: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Self advertisement

For N = 2, there are solutions to restricted models of boundedmemory.

- Neyman (2008) and Peretz (forthcoming MOR) solveG (moblivious , 2

Θm),

- Peretz (2011 GEB) solves G (krecall ,mrecall).

For N = 3,

- Peretz (forthcoming IJGT) suggests some bounds for the mixedmin-max of G (Θmrecall ,mrecall ,mrecall).

- The present work facilitates bounds for the mixed max-min ofG (2Θm logm,m,m) and pure min-max of G (Θm,m,m).

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 6 / 14

Page 50: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Plan

.Key question..

......

What kind of probability distributions on plays can result from probabilitydistributions on pairs of m-memory strategies?

Let N = 2. The set of m-memory strategies of player i is denotedAi (m).

A pair (σ, τ) ∈ A1(m)×A2(m) generates a play x1, x2, . . .

Not every play can be generated this way. For example, the play mustenter a loop in the first m2 periods.

A probability distribution µ ∈ ∆(A1(m)×A2(m)) induces aprobability distribution on plays, µ ∈ ∆(AN).

What is the range of µ?

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 7 / 14

Page 51: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Plan

.Key question..

......

What kind of probability distributions on plays can result from probabilitydistributions on pairs of m-memory strategies?

Let N = 2. The set of m-memory strategies of player i is denotedAi (m).

A pair (σ, τ) ∈ A1(m)×A2(m) generates a play x1, x2, . . .

Not every play can be generated this way. For example, the play mustenter a loop in the first m2 periods.

A probability distribution µ ∈ ∆(A1(m)×A2(m)) induces aprobability distribution on plays, µ ∈ ∆(AN).

What is the range of µ?

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 7 / 14

Page 52: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Plan

.Key question..

......

What kind of probability distributions on plays can result from probabilitydistributions on pairs of m-memory strategies?

Let N = 2. The set of m-memory strategies of player i is denotedAi (m).

A pair (σ, τ) ∈ A1(m)×A2(m) generates a play x1, x2, . . .

Not every play can be generated this way. For example, the play mustenter a loop in the first m2 periods.

A probability distribution µ ∈ ∆(A1(m)×A2(m)) induces aprobability distribution on plays, µ ∈ ∆(AN).

What is the range of µ?

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 7 / 14

Page 53: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Plan

.Key question..

......

What kind of probability distributions on plays can result from probabilitydistributions on pairs of m-memory strategies?

Let N = 2. The set of m-memory strategies of player i is denotedAi (m).

A pair (σ, τ) ∈ A1(m)×A2(m) generates a play x1, x2, . . .

Not every play can be generated this way. For example, the play mustenter a loop in the first m2 periods.

A probability distribution µ ∈ ∆(A1(m)×A2(m)) induces aprobability distribution on plays, µ ∈ ∆(AN).

What is the range of µ?

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 7 / 14

Page 54: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Plan

.Key question..

......

What kind of probability distributions on plays can result from probabilitydistributions on pairs of m-memory strategies?

Let N = 2. The set of m-memory strategies of player i is denotedAi (m).

A pair (σ, τ) ∈ A1(m)×A2(m) generates a play x1, x2, . . .

Not every play can be generated this way. For example, the play mustenter a loop in the first m2 periods.

A probability distribution µ ∈ ∆(A1(m)×A2(m)) induces aprobability distribution on plays, µ ∈ ∆(AN).

What is the range of µ?

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 7 / 14

Page 55: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Plan

.Key question..

......

What kind of probability distributions on plays can result from probabilitydistributions on pairs of m-memory strategies?

Let N = 2. The set of m-memory strategies of player i is denotedAi (m).

A pair (σ, τ) ∈ A1(m)×A2(m) generates a play x1, x2, . . .

Not every play can be generated this way. For example, the play mustenter a loop in the first m2 periods.

A probability distribution µ ∈ ∆(A1(m)×A2(m)) induces aprobability distribution on plays, µ ∈ ∆(AN).

What is the range of µ?

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 7 / 14

Page 56: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Approximation of i.i.d. random variables

.Lemma..

......

For every ϵ > 0 there exists δ > 0 such that for every Q ∈ ∆(A), n ∈ Nand random variables x1, . . . , xn assuming values in A, if

...1 ∥E [emp(x1, . . . , xn)]− Q∥ < δ, and

...2 1nH(x1 . . . , xn) > H(Q)− δ,

then1

n

n∑k=1

E ∥Pr(xk |x1, . . . , xk−1)− Q∥ < ϵ,

and vice versa.

Where H is Shanon’s entropy function and emp is the empirical frequency,the number of times each letter appears in the sequence divided by n.

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 8 / 14

Page 57: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Implementation through bounded memory strategies

For a given Q ∈ ∆(A) we would like to find the largest n = n(m,Q,A)such that one can approximate n i.i.d. random variables with distributionQ through pairs of m-memory strategies..Defenition..

......

C (Q) = inf{C > 0 : ∃nm ∈ N, µm ∈ ∆(A1(m)×A2(m)) such that

C · nm ≥ m logm, and

the induced play xm1 , . . . , xmnm satisfies

(1) ∥E[emp(xm1 , . . . , xmnm)]− Q∥ < o(1),

(2) 1/nm H(xm1 , . . . , xmnm) > H(Q)− o(1).}

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 9 / 14

Page 58: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Main result

.Theorem..

......max

i

(H(Qi )

|A−i | − 1

)≤ C (Q) ≤ max

i

(H(Q)

|Ai | − 1

)

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 10 / 14

Page 59: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Proof: lower bound

Define

Ci (Q) = inf{C : ∃nm ∈ N, µm ∈ ∆(Ai (m)× Anm

−i ) such that

C · nm ≥ m logm, and

the induced play xm1 , . . . , xmnm satisfies

(1) ∥E[emp(xm1 , . . . , xmnm)]− Q∥ < o(1),

(2)1

nmH(xm1 , . . . , xmnm) > H(Q)− o(1).

}We show that

H(Qi )

|A−i | − 1≤ Ci (Q).

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 11 / 14

Page 60: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Proof “H(Qi) ≤ (|A−i | − 1)Ci(Q)”

Let C > 0 such that there exists n ≥ m logm/C and µ ∈ ∆(Ai (m)× An−i )

that induces a play that satisfies...1 ∥E[emp(x1, y1, . . . , xn, yn)]− Q∥ < o(1),...2 1

nH(x1, y1, . . . , xn, yn) > H(Q)− o(1).

Let σ and Y be random variables that distribute according to µ. We have

H(Qi ) + HQ(−i |i)− o(1) = H(Q)− o(1) ≤1

nH(x1, . . . , yn) ≤

1

nH(σ,Y ) =

1

nH(σ) +

1

nH(Y |σ)

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 12 / 14

Page 61: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Proof “H(Qi) ≤ (|A−i | − 1)Ci(Q)”

Let C > 0 such that there exists n ≥ m logm/C and µ ∈ ∆(Ai (m)× An−i )

that induces a play that satisfies...1 ∥E[emp(x1, y1, . . . , xn, yn)]− Q∥ < o(1),...2 1

nH(x1, y1, . . . , xn, yn) > H(Q)− o(1).

Let σ and Y be random variables that distribute according to µ.

We have

H(Qi ) + HQ(−i |i)− o(1) = H(Q)− o(1) ≤1

nH(x1, . . . , yn) ≤

1

nH(σ,Y ) =

1

nH(σ) +

1

nH(Y |σ)

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 12 / 14

Page 62: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Proof “H(Qi) ≤ (|A−i | − 1)Ci(Q)”

Let C > 0 such that there exists n ≥ m logm/C and µ ∈ ∆(Ai (m)× An−i )

that induces a play that satisfies...1 ∥E[emp(x1, y1, . . . , xn, yn)]− Q∥ < o(1),...2 1

nH(x1, y1, . . . , xn, yn) > H(Q)− o(1).

Let σ and Y be random variables that distribute according to µ. We have

H(Qi ) + HQ(−i |i)− o(1) = H(Q)− o(1) ≤

1

nH(x1, . . . , yn) ≤

1

nH(σ,Y )

=1

nH(σ) +

1

nH(Y |σ)

.Explanation........ The play is a function of σ and Y .

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 12 / 14

Page 63: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Proof “H(Qi) ≤ (|A−i | − 1)Ci(Q)”

Let C > 0 such that there exists n ≥ m logm/C and µ ∈ ∆(Ai (m)× An−i )

that induces a play that satisfies...1 ∥E[emp(x1, y1, . . . , xn, yn)]− Q∥ < o(1),...2 1

nH(x1, y1, . . . , xn, yn) > H(Q)− o(1).

Let σ and Y be random variables that distribute according to µ. We have

H(Qi ) + HQ(−i |i)− o(1) =

H(Q)− o(1) ≤1

nH(x1, . . . , yn) ≤

1

nH(σ,Y )

=1

nH(σ) +

1

nH(Y |σ)

.Explanation........

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 12 / 14

Page 64: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Proof “H(Qi) ≤ (|A−i | − 1)Ci(Q)”

Let C > 0 such that there exists n ≥ m logm/C and µ ∈ ∆(Ai (m)× An−i )

that induces a play that satisfies...1 ∥E[emp(x1, y1, . . . , xn, yn)]− Q∥ < o(1),...2 1

nH(x1, y1, . . . , xn, yn) > H(Q)− o(1).

Let σ and Y be random variables that distribute according to µ. We have

H(Qi ) + HQ(−i |i)− o(1) =

H(Q)− o(1) ≤1

nH(x1, . . . , yn) ≤

1

nH(σ,Y ) =

1

nH(σ) +

1

nH(Y |σ)

.Explanation........ Chain rule.

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 12 / 14

Page 65: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Proof “H(Qi) ≤ (|A−i | − 1)Ci(Q)”

Let C > 0 such that there exists n ≥ m logm/C and µ ∈ ∆(Ai (m)× An−i )

that induces a play that satisfies...1 ∥E[emp(x1, y1, . . . , xn, yn)]− Q∥ < o(1),...2 1

nH(x1, y1, . . . , xn, yn) > H(Q)− o(1).

Let σ and Y be random variables that distribute according to µ. We have

H(Qi ) + HQ(−i |i)− o(1) = H(Q)− o(1) ≤1

nH(x1, . . . , yn) ≤

1

nH(σ,Y ) =

1

nH(σ) +

1

nH(Y |σ)

.Explanation........ Chain rule.

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 12 / 14

Page 66: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Proof “H(Qi) ≤ (|A−i | − 1)Ci(Q)”

Let C > 0 such that there exists n ≥ m logm/C and µ ∈ ∆(Ai (m)× An−i )

that induces a play that satisfies...1 ∥E[emp(x1, y1, . . . , xn, yn)]− Q∥ < o(1),...2 1

nH(x1, y1, . . . , xn, yn) > H(Q)− o(1).

Let σ and Y be random variables that distribute according to µ. We have

H(Qi ) + HQ(−i |i)− o(1) = H(Q)− o(1) ≤1

nH(x1, . . . , yn) ≤

1

nH(σ,Y ) =

1

nH(σ) +

1

nH(Y |σ)

.Explanation........ Suppressing the middle part...

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 12 / 14

Page 67: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Proof “H(Qi) ≤ (|A−i | − 1)Ci(Q)”

Let C > 0 such that there exists n ≥ m logm/C and µ ∈ ∆(Ai (m)× An−i )

that induces a play that satisfies...1 ∥E[emp(x1, y1, . . . , xn, yn)]− Q∥ < o(1),...2 1

nH(x1, y1, . . . , xn, yn) > H(Q)− o(1).

Let σ and Y be random variables that distribute according to µ. We have

H(Qi ) + HQ(−i |i)− o(1) ≤ 1

nH(σ) +

1

nH(Y |σ)

1

nH(σ) + HE[emp(x1,...,yn)](−i |i) ≤ C · log |Ai (m)|

m logm

.Explanation........

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 12 / 14

Page 68: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Proof “H(Qi) ≤ (|A−i | − 1)Ci(Q)”

Let C > 0 such that there exists n ≥ m logm/C and µ ∈ ∆(Ai (m)× An−i )

that induces a play that satisfies...1 ∥E[emp(x1, y1, . . . , xn, yn)]− Q∥ < o(1),...2 1

nH(x1, y1, . . . , xn, yn) > H(Q)− o(1).

Let σ and Y be random variables that distribute according to µ. We have

H(Qi ) + HQ(−i |i)− o(1) ≤ 1

nH(σ) +

1

nH(Y |σ) ≤

1

nH(σ) + HE[emp(x1,...,yn)](−i |i)

≤ C · log |Ai (m)|m logm

.Explanation........ Concavity, Neyman-Okada (2009).

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 12 / 14

Page 69: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Proof “H(Qi) ≤ (|A−i | − 1)Ci(Q)”

Let C > 0 such that there exists n ≥ m logm/C and µ ∈ ∆(Ai (m)× An−i )

that induces a play that satisfies...1 ∥E[emp(x1, y1, . . . , xn, yn)]− Q∥ < o(1),...2 1

nH(x1, y1, . . . , xn, yn) > H(Q)− o(1).

Let σ and Y be random variables that distribute according to µ. We have

H(Qi ) + HQ(−i |i)− o(1) ≤ 1

nH(σ) +

1

nH(Y |σ) ≤

1

nH(σ) + HE[emp(x1,...,yn)](−i |i)

≤ C · log |Ai (m)|m logm

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 12 / 14

Page 70: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Proof “H(Qi) ≤ (|A−i | − 1)Ci(Q)”

Let C > 0 such that there exists n ≥ m logm/C and µ ∈ ∆(Ai (m)× An−i )

that induces a play that satisfies...1 ∥E[emp(x1, y1, . . . , xn, yn)]− Q∥ < o(1),...2 1

nH(x1, y1, . . . , xn, yn) > H(Q)− o(1).

Let σ and Y be random variables that distribute according to µ. We have

H(Qi ) − o(1) ≤ 1

nH(σ)

≤ C · log |Ai (m)|m logm

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 12 / 14

Page 71: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Proof “H(Qi) ≤ (|A−i | − 1)Ci(Q)”

Let C > 0 such that there exists n ≥ m logm/C and µ ∈ ∆(Ai (m)× An−i )

that induces a play that satisfies...1 ∥E[emp(x1, y1, . . . , xn, yn)]− Q∥ < o(1),...2 1

nH(x1, y1, . . . , xn, yn) > H(Q)− o(1).

Let σ and Y be random variables that distribute according to µ. We have

H(Qi ) − o(1) ≤ 1

nH(σ) ≤ C · log |Ai (m)|

m logm

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 12 / 14

Page 72: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Proof “H(Qi) ≤ (|A−i | − 1)Ci(Q)”

Let C > 0 such that there exists n ≥ m logm/C and µ ∈ ∆(Ai (m)× An−i )

that induces a play that satisfies...1 ∥E[emp(x1, y1, . . . , xn, yn)]− Q∥ < o(1),...2 1

nH(x1, y1, . . . , xn, yn) > H(Q)− o(1).

Let σ and Y be random variables that distribute according to µ. We have

H(Qi ) − o(1) ≤ 1

nH(σ) ≤ C · log |Ai (m)|

m logm

It remains to show that log |Ai (m)|/m logm ≤ |A−i | − 1 + o(1).

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 12 / 14

Page 73: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Proof: “log |Ai(m)|/m logm ≤ |A−i | − 1 + o(1)”

In order to estimate |Ai (m)| from above we use a combinatorialstructure to describe finite memory strategies.

An m-memory strategy for player i can be executed by a finiteautomaton with m-states, input alphabet A−i and output alphabet Ai .

Counting these automata and dividing by m! (for renaming of states)shows that log |Ai (m)|/m logm ≤ |A−i − 1|+ o(1).

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 13 / 14

Page 74: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Lemma (Neyman-Okada 2009)

Let X and τ be random variables that take values

X in An1,

τ in the set of pure strategies of player 2 in the n-fold repeated game.

The pair (X , τ) generate a play (x1, y1, . . . , xn, yn), where X = (x1, . . . , xn)and yt = τ(x1, y1 . . . , xt−1, yt−1). Let a1 and a2 be random variableswhose joint distribution is the expected empirical frequency of the inducedplay. We have

H(a1|a2) ≥1

nH(X |τ).

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 14 / 14

Page 75: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Lemma (Neyman-Okada 2009)

Let X and τ be random variables that take values

X in An1,

τ in the set of pure strategies of player 2 in the n-fold repeated game.

The pair (X , τ) generate a play (x1, y1, . . . , xn, yn), where X = (x1, . . . , xn)and yt = τ(x1, y1 . . . , xt−1, yt−1).

Let a1 and a2 be random variableswhose joint distribution is the expected empirical frequency of the inducedplay. We have

H(a1|a2) ≥1

nH(X |τ).

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 14 / 14

Page 76: Correlation Through Bounded Memory Strategies...Correlation Through Bounded Memory Strategies Ron Peretz1 (joint work with Olivier Gossner and Pen´elope Hern´andez) 1Tel Aviv University

. . . . . .

Lemma (Neyman-Okada 2009)

Let X and τ be random variables that take values

X in An1,

τ in the set of pure strategies of player 2 in the n-fold repeated game.

The pair (X , τ) generate a play (x1, y1, . . . , xn, yn), where X = (x1, . . . , xn)and yt = τ(x1, y1 . . . , xt−1, yt−1). Let a1 and a2 be random variableswhose joint distribution is the expected empirical frequency of the inducedplay. We have

H(a1|a2) ≥1

nH(X |τ).

Ron Peretz ([email protected]) Bounded Memory Correlation November, 2011 14 / 14