Consider an infinitely repeated prisoner's dilemma game by two players. The resultant payoffs at each stage by the actions of two players are given below in the table (payoffs are denoted like (payoff for player 1, payoff for player 2)). Two players determine their strategies simultaneously and independently to maximize expected payoffs of their own based on their information. The game is potentially infinitely repeated, however, the game ends at a probability of 1-x (0≤1-x≤1) in every stage (that is, these players continue to play this game at the probability of x). There is no discount rate for future payoffs (i.e. both players weight current and future payoffs equally).
|
|
player 2
|
|
C
|
D
|
|
player 1
|
C
|
(9,9)
|
(1,13)
|
|
D
|
(13,1)
|
(3,3)
|
(a) Suppose two players adopt a Trigger Strategy (Play C in the first stage. In the tth stage (t≥2), if the outcome of all t-1 preceding stages has been (9, 9), then play C; otherwise, play D). Find the range of x which makes cooperation self-sustainable.
(b) Suppose two players adopt a Tit for Tat Strategy (TFT) (Play C in the first stage. And then, do whatever the other player did at the previous stage). Find the range of x which makes cooperation self-sustainable.