Modelling individual behaviour and group performance in an intelligent agent-based simulation of the tragedy of the commons

This work introduces and illustrates the potential of intelligent agent-based modelling and simulation as a tool for understanding individual action and group performance in common-pool resource dilemmas. A model was developed, based on previously documented common-pool resource experiments, and simulated using the Swarm multi-agent simulation environment. Agents in these models were designed to represent the actions of the individual appropriators in the experiments and the common-pool resource itself. The intelligent agents developed for these simulations possess adaptive capabilities. They select amongst alternative strategies the one that appears to yield the best performance under current conditions. Simulations of this model rendered observations of some potential relationships between individual action and group performance in common-pool resource experimental situations. In particular, adaptive agents generated group level behaviour with similar performance characteristics to groups in actual experiments. Some potential future directions for this research, and possible applications in natural resources management, are discussed.