Combining Influence Maps with Heuristic Search for Executing Sneak-Attacks in RTS Games

Abstract

Real-Time Strategy (RTS) games have become a popular domain for AI research due to their large state and action spaces, as well as complex sub-problems. One popular strategy in RTS games is the idea of a Sneak-Attack, in which one player attempts to sneak enemy units into the base of their opponent without being seen, in order to gain the element of surprise. In this paper we will present initial results on combining influence maps with heuristic search to produce a path-finding system which allows us to guide StarCraft drop ships in order to execute a sneak attack. Our preliminary results show that by combining these two techniques, we can efficiently and automatically produce paths that guide our drop ships in a stealthy manner toward the enemy base, minimizing distance traveled and avoiding enemy vision of our army.