Lecture 14 Infinitely Repeated Games II 14.12 Game Theory Muhamet - - PDF document

lecture 14
SMART_READER_LITE
LIVE PREVIEW

Lecture 14 Infinitely Repeated Games II 14.12 Game Theory Muhamet - - PDF document

Lecture 14 Infinitely Repeated Games II 14.12 Game Theory Muhamet Yildiz 1 Road Map 1. Folk Theorem 2. Applications (Problems) 2 ~. Folk Theorem Definition: v = (v 1 ,v , . .. ,vn) is feasible iff v is a convex 2 combination of pure-strategy


slide-1
SLIDE 1

Lecture 14 Infinitely Repeated Games II

14.12 Game Theory Muhamet Yildiz

1

slide-2
SLIDE 2

Road Map

  • 1. Folk Theorem
  • 2. Applications (Problems)

2

slide-3
SLIDE 3

Folk Theorem

Definition: v = (v1,v , . ..

2

,vn) is feasible iff v is a convex

combination of pure-strategy payoff-vectors: 1 2 m v = p1 u(a ) + p2u(a ) + ... + Pnu(a ), where P1 + P2 + ... + Pm = 1, and u(ai) is the payoff vector at strategy profile ai of the stage game. Theorem: Let x = (X ,X , ...

) 1 2

,xn be s feasible payoff vector, and e = (e ,e , . . .

1 2

,en) be a payoff vector at some equilibrium of the stage game such that Xi >

ei for each i. Then, there exist ~ < 1 and a strategy

profile s such that s yields x as the expected average-payoff vector and is a SPE whenever 8 >

~.

3

slide-4
SLIDE 4

Folk Theorem in PO

C D

  • A SPE with PV

C 5,5 0,6

(1.1,1.1)?

D 6,0 1,1

  • With PV (1 .1

,5)?

  • With PV (6,0)?
  • With PV (5.9,0.1)?

4

slide-5
SLIDE 5

Proof for a special case

  • Assume x = u(a*) = (u1(a*), ...

, un(a*» for some a*.

  • s*: Every player i plays ai* until somebody

deviates and plays ei thereafter.

  • Average value of

i from s* is Xi = ul

a

*).

  • s* is a SPE ¢:> {) > 8 where

5

slide-6
SLIDE 6

Applications/Problems

6

slide-7
SLIDE 7

2010 Midterm 2, P2

2.

  • 1
  • l
  • t.

2

  • 1

3

  • ;L

:;

  • i

D

7

slide-8
SLIDE 8

Range of 8 for SPE

  • Alice Hires and Bob and Colin both Work

until any of the workers Shirk; Alice Hires and Bob and Colin both Shirk thereafter.

  • Alice Always Hires. Both workers Work at

t = 0. At any t > 0, each worker Works if the previous play is (Hire, Work, Work) or (Hire, Shirk, Shirk); each worker Shirks

  • therwise.

8

slide-9
SLIDE 9

2007 Midterm 2, P3

  • Stage Game: Linear Bertrand Duopoly (c=O; Q=I-p)
  • s*: They both charge 112 until somebody deviates;

they both charge 0 thereafter.

  • s**: n + I modes: Collusion, WI, W2, ... , Wn. Game

starts at Collusion. Both charge 112 in the Collusion mode and p*<112 in WI,

... , Wn. Without deviation,

Collusion leads to Collusion, WI leads to W2, ... , Wn-I leads to Wn, and Wn leads to Collusion. Any deviation leads to WI.

9

slide-10
SLIDE 10

MIT OpenCourseWare http://ocw.mit.edu

14.12 Economic Applications of Game Theory

Fall 2012 For information about citing these materials or our Terms of Use, visit: http://ocw.mit.edu/terms.