An experimental analysis of Colonel Blotto Games under alternative environments

In this paper, we design an experiment based on the Colonel Blotto game with complete information under auction method in which the player allocating more resources to a battlefield wins with certainty. The main contribution of this paper is that we allow the same player to play sequentially under different environments (symmetric and asymmetric resources and different number of battlefields) against computer. Consistent with the theoretical predictions, we find that most of the time players use a ‘complete coverage’ strategy when they have equal or more resources than the opponent and use a ‘guerilla warfare’ strategy when they have fewer resources. We also find that an individual player’s success is relatively high when the player has equal or more resources than the opponent. Finally, we observe that there is an improvement in players’ success both within a treatment and across similar treatments which can be explained by “learning” argument.