Intelligent Battle Gaming Pragmatics with Belief Network Trees

The events in an evolving battlespace unfold under the partial control of adversarial commanders, who each attempt to force the situation into a favorable state according to their goals. The player here must match wits with a bot by using available resources and situation models to achieve a goal. Unlike checkers or chess, or characters acting out scripts, the current situations are not known with certainty but can only be estimated from associations and correlations with other uncertain data. Further, either side can make multiple moves without waiting for the other side to act, so timeliness is critical. The player selects a scenario, weapons, a goal and parameters to initialize such games. The adversarial bot starts with a set of rules that expand by experience using case-based reasoning. The philosophical and pragmatic approach taken here lays out a scheme for battle gaming, whether real or recreational, by means of a tree of belief networks to aid in the decision making at each move. This preliminary investigation precedes the future development of algorithms