Abstract: A* is a widely used algorithm for path-finding in the area of Artificial Intelligence for video games. Even though this algorithm guarantees optimality, it tends to bring forth similar behaviors in agents that are spatially close to each other. On the other hand, when the agents are sparsely distributed, the algorithm doesn't ensure an attack that comes from different places. We propose four variations of A* that produce ambush behaviors and diversity of paths: A*mbush, P-A*mbush, R-A*mbush and SAR-A*mbush. They are modifications of A* that take into account the number of agents that have a specific node (or edge) in their calculated path in order to vary the cost function of the graph. P-A*mbush, R-A*mbush and SAR-A*mbush are variations of A*mbush that improve the paths generated by the proposed A*mbush algorithm.