Simulations of assortative matching: complexity of cooperation emergence process

Abstract

The paper studies the emergence and stability of cooperative behavior in populations of agents who interact among themselves in Prisoner's Dilemma games and who are allowed to choose their partners. The population is then subject to evolutionary pressures, based on individual payoffs from multiple encounters. A simple formula for signaling and recognition is introduced, which allows the agents to guess prospective partner strategy and to refuse the game if the predicted outcome is unfavorable. We present both algebraic formulation of the average payoffs and results of computer simulations of evolution of such society. The simulations result in surprising variety of behavior. We discuss possible interpretations of the results as well as relationship between artificial `computer time' and real time of possible social and biological systems that the model might be applied to.