University of AlbertaFrantisek Sailer
RTS game developers face the problem of creating AI systems capable of playing well in a team and against a large variety of opponents. The common approach is to script behaviors with the goal of covering the most important strategies used by human players. Creating scripts that perform well for RTS games, however, is hard— it resembles writing good chess programs based solely on static rules without a way of evaluating future developments. To offset the lack of such planning capabilities, game designers often resort to giving AI systems unfair advantages, such as complete ...