General Blotto: games of allocative strategic mismatch

Abstract

The Colonel Blotto game captures strategic situations in which players attempt to mismatch an opponent’s action. We extend Colonel Blotto to a class of General Blotto games that allow for more general payoffs and externalities between fronts. These extensions make Blotto applicable to a variety of real-world problems. We find that like Colonel Blotto, most General Blotto games do not have pure strategy equilibria. Using a replicator dynamics learning model, we show that General Blotto may have more predictable dynamics than the original Blotto game. Thus, adding realistic structure to Colonel Blotto may, paradoxically, make it less complex.