Artificial Ants and Dynamical Adaptation of Accessible Games Level

Designing a relevant artificial intelligence engine for video games does not always consist in finding a best solution (best opponent, best path, etc.), it can sometimes consist in providing the player with the best gaming experience, and this can be a compromise between being strong enough to give an interesting challenge, and not being too strong so that the player can never win. This game tuning is particularly prominent in case of games for impaired children: they need to be encouraged more than others, and their tiredness can occur quickly within a game that does not fit to their level. In this paper, we focus on self adaptation abilities of ant colonies in order to test if this kind of behavior can be an efficient way to deal with the game level adaptation problem.