Feasibility of Applying a Genetic Algorithm to Playing Stratego

Abstract

In this thesis, I investigate whether applying genetic algorithms to playing Stratego is likely to produce useful results, i.e. artificial players that can match (skilled) human players. Applying a simple parameter-based genetic algorithm resulted in a noticeable improvement in play that could match other artificial players. I end with a number of suggestions on how to improve on this algorithm with more sophisticated techniques to drastically improve performance so as to be able to actually be challenging for human players to play against.