Cooperation under Incomplete Information on the Discount Factors
📝 Original Info
- Title: Cooperation under Incomplete Information on the Discount Factors
- ArXiv ID: 1012.3117
- Date: 2015-09-03
- Authors: Researchers from original ArXiv paper
📝 Abstract
In the repeated Prisoner's Dilemma, when every player has a different discount factor, the grim-trigger strategy is an equilibrium if and only if the discount factor of each player is higher than some threshold. What happens if the players have incomplete information regarding the discount factors? In this work we look at repeated games in which each player has incomplete information regarding the other player's discount factor, and ask when a pair of grim-trigger strategies is an equilibrium. We provide necessary and sufficient conditions for such strategies to be an equilibrium. We characterize the states of the world in which the strategies are not triggered, i.e., the players cooperate, in such equilibria (or $\epsilon$-equilibria), and ask whether these "cooperation events" are close to those in the complete information case, when the information is "almost" complete, in several senses.💡 Deep Analysis
Deep Dive into Cooperation under Incomplete Information on the Discount Factors.In the repeated Prisoner’s Dilemma, when every player has a different discount factor, the grim-trigger strategy is an equilibrium if and only if the discount factor of each player is higher than some threshold. What happens if the players have incomplete information regarding the discount factors? In this work we look at repeated games in which each player has incomplete information regarding the other player’s discount factor, and ask when a pair of grim-trigger strategies is an equilibrium. We provide necessary and sufficient conditions for such strategies to be an equilibrium. We characterize the states of the world in which the strategies are not triggered, i.e., the players cooperate, in such equilibria (or $\epsilon$-equilibria), and ask whether these “cooperation events” are close to those in the complete information case, when the information is “almost” complete, in several senses.