Cooperative Artificial Intelligence In The Euchre Card Game

Abstract

In order to provide a method for optimization in multi-agent systems using artificial intelligence (AI), a set of logic rules provides information for traversing options and maximizing expected utilities. This project explores the use of logical decision making in multi-player games, specifically in the card game Euchre. A variety of algorithms implemented herein compare decisions as quickly as possible while still trying to optimize their moves. Based on these comparisons, a better method for optimizing strategies becomes evident. This implementation contains seven AI agents with some slight variations on those AI agents. Two pairs of AI agents will work together against another pair of agents in order to maximize not only their own personal goals, but their team score. The complexity of these agents having to work on a solitary level as well as a team level takes AI to a powerful level. This thesis also presents some promising results, compared to some other Euchre programs, for how AI can cooperate as a team, whether or not the AI type is the same or different.