Reconstruction of micro-dynamics characterizing human decision-making behavior in repeated social dilemmas

Abstract

Exploring evolutionary updating rules more consistent with individual cognitive processes is crucial to the study of human cooperation. A considerable number of dynamic models describing human decision-making behavior lack empirical evidence. We have conducted a behavioral experiment and proposed a hypothesis that human players make decisions based on proportional change rather than absolute difference of payoffs. Thus we can reconstruct the micro-dynamics, namely individual strategy updating rules, with a singular function considering the individual's choice to cooperate and defect separately. We find the micro-dynamics evolve over time and our dynamic model can well predict the evolution of cooperation. Furthermore, combined with the conception “elasticity” commonly used in economics, we reveal the relationship between the intrinsic property of population and players’ subjective decision intentions. Our findings provide a novel framework for measuring human decision-making behavior with bounded rationality in repeated social dilemmas.