A non-equilibrium analysis of the finitely-repeated prisoner's dilemma (Q1108949)

We follow a non-equilibrium approach to the finitely repeated prisoner's dilemma with trigger strategies. Each player has a probability distribution which gives the probability with which he thinks the other player first plans to cheat in any given period. We show that, provided that this probability distribution assigns some minimum weight to all periods, the players will cooperate for most of the game if it is repeated enough times.