On the human-driven decision-making process in competitive environments

The emergence of intelligent sensing and communication technologies fosters the generation and dissemination of huge amounts of information that collectively enriches people’s awareness about their environment and its resources. With this information at hand, users then decide how to access these resources to best serve their interests. However, situations repeatedly emerge where the users’ welfare is better satisfied by the same finite set of resources and the uncoordinated access to them gives rise to tragedy of commons effects and serious congestion problems. In this paper, we study generic scenarios, where some nonexcludable finite resource is of interest to a population of distributed users with variable perceptions about the resource supply and demand for it. The high-level question we address is how efficiently the competition about the resources is resolved under different assumptions about the way the users make their decisions. The users are first viewed as strategic perfectly informed software agents that make fully rational decisions attempting to minimize the cost of accessing the acquired resource. We then exploit insights from experimental economics and cognitive psychology to model agents of bounded rationality who either do not possess perfect information or cannot exploit all the available information due to time restrictions and computational limitations. We derive the operational states in which the competing influences are balanced (i.e., equilibria) and compare them against the Nash equilibria that emerge under full rationality and the optimum resource assignment that could be determined by a centralized entity. Our results provide useful insights to the dynamics emerging from the agents’ behavior as well as theoretical support for the practical management of limitedcapacity resources.