Learning about learning in games through experimental control of strategic interdependence.

Shachat, J.; Swarthout, J.T.

Abstract

We report results from an experiment in which humans repeatedly play one of two games against a computer program that follows either a reinforcement or an experience weighted attraction learning algorithm. Our experiment shows 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 does not vary significantly. These factors lead to a strong linear relationship between the humans' and algorithms' action choice proportions that is suggestive of the algorithms' best response correspondences.

• Journal Article Type: Article
• Publication Date: 2012-03
• Deposit Date: Sep 17, 2014
• Journal: Journal of Economic Dynamics and Control
• Print ISSN: 0165-1889
• Publisher: Elsevier
• Peer Reviewed: Peer Reviewed
• Volume: 36
• Issue: 3
• Pages: 383-402
• DOI: https://doi.org/10.1016/j.jedc.2011.09.007
• Public URL: https://durham-repository.worktribe.com/output/1420843