抄録
In this paper, by simulation of an agent-based model, we analyze cooperative behavior of the two players in the prisoner’s dilemma game in which each of the two players has two strategies: ‘cooperate’ and ‘defect’. The strategy ‘defect’ is rational one for both players in the repeated prisoner’s dilemma games with finite sequence of rounds. However, in the games with infinite sequence of rounds, it is known that a pair of strategies that both of the players choose strategy ‘cooperate’ can be a Nash equilibrium solution. To verify such cooperative behavior, we employ the Sugarscape model which is an artificial society model developed by Epstein and Axtell, and examine the repeated prisoner’s dilemma game in which decisions of the agents are randomly made on the basis of history of the past three decisions. For this simulation, we trace agents to survive and investigate what cooperative behavior emerges and grows in the model.
