How bounded rationality of individuals in social interactions impacts evolutionary dynamics of cooperation

Abstract

In this study, we explore the emergence of cooperative behavior in the prisoner's dilemma evolutionary game. In particular, we investigate the effect of bounded rationality of individuals on the networking topology (i.e., the individuals' personal networks). For this, we highlight the evolutionary dynamics of cooperation on top of different graph topologies with respect to their baseline properties such as average shortest path length and clustering coefficient. In addition, we test the effect of a new variable, called memory of interactions, on the changes in behavior and decision-making of the players as well as the networking outcome. For this purpose, we use agent-based modeling, which allows studying how changes in the environment or changes of properties of networked actors affect the evolutionary dynamics of cooperation among them. The results of our analysis confirm that the networking topology and the memory duration are important in affecting the emergence of cooperative behavior of players. They also impact the total utility that can be obtained from playing the Prisoner's Dilemma evolutionary game. Although the Prisoner's Dilemma game simulations tend towards full cooperation, if they are run over graph topologies with short average shortest path lengths and low clustering coefficients, the number of steps needed to reach equilibrium increases. This new result provides an understanding of the interactions of actors in a game.