Learning About Learning in Games Through Experimental Control of Strategic Interdependence

Abstract

We report experiments in which humans repeatedly play one of two games against a computer program that follows either a reinforcement learning or an Experience Weighted Attraction algorithm. Our experiments show these learning algorithms detect exploitable opportunities more sensitively than humans. Also, learning algorithms respond to detected payoff-increasing opportunities systematically; however, the responses are too weak to improve the algorithms' payoffs. Human play against various decision maker types doesn't vary significantly. These factors lead to a strong linear relationship between the humans' and algorithms' action choice proportions that is suggestive of the algorithm's best response correspondence. These properties are revealed only by our human versus computer experiments, and not by our standard human versus human experiments, nor our model simulations.

Keywords: learning, repeated games, experiments, simulation

JEL Classification: C72, C92, C81