An initial implementation of the Turing tournament to learning in repeated two-person games

We report on a design of a Turing tournament and its initial implementation to learning in repeated 2-person games. The principal objectives of the tournament, named after the original Turing Test, are (1) to find teaming algorithms (emulators) that most closely simulate human behavior, (2) to find algorithms (detectors) that most accurately distinguish between humans and machines, and (3) to provide a demonstration of how to implement this methodology for evaluating models of human behavior. In order to test our concept, we developed the software and implemented a number of learning models well known in the literature and developed a few detectors. This initial implementation found significant differences in data generated by these learning models and humans, with the greatest ones in coordination games. Finally, we investigate the stability of our result with respect to different evaluation approaches. (c) 2006 Elsevier Inc. All rights reserved.