Emergence of cooperator dominance in an iterated prisoner's dilemma games with punishing agents and dynamic strategies

Abstract

The evolution of cooperation in modern societies may be attributed to the central law-enforcing body, and not to individual retaliation. In this study, aside from the classical players of the Prisoner's dilemma game (cooperating C and defecting D agents), we introduced the punishing agents which represents the law-enforcing body, which we refer to the as Prisoner's Dilemma-Punisher method (PD-P). In-silico based experiments kept the frequency of these agents constant while allowing for probabilistic changes in strategy of the playing agents C and D. Strategy changes include mutation where addition of the punishing agents results to a higher cooperator density than when there were no punishing agents. Another method was the strategy imitation of the best performing neighbor where it resulted to a higher dierence in the density of the cooperators and defectors than when there were no punishing agents in the lattice. From the results of this experiment, we show how the addition of the punishing agents, together with allowing probabilistic changes in strategy results to an emergence of a generally more cooperative community.