Using ants' task division for better game engines - A contribution to game accessibility for impaired players

Abstract

Designing a relevant artificial intelligence engine for video games does not always consist in finding the best solution (best opponent, best path, etc.): it can sometimes consist in offering the player the best gaming experience. Such a good experience is linked with the difficulty level of the proposed challenge. A game that is too easy will be boring whereas a game that is too difficult will be stressful. So, to be interesting for everyone, an artificial intelligence engine should provide an adaptive game level for every player. This game tuning is particularly prominent in the especial case of accessible games for impaired player. In this paper we show that the task division model based on ant colonies can be an interesting way to provide adaptive behaviors in game engines for simple one-player games. © 2008 Springer-Verlag Berlin Heidelberg.