An experimental investigation of colonel blotto games

This article examines behavior in the Colonel Blotto game with asymmetric forces. In this constant-sum game, two players simultaneously allocate their forces across n-battlefields, with the objective of maximizing the expected number of battlefields won. The experimental results support all major theoretical predictions. In the auction treatment, where winning a battlefield is deterministic, the disadvantaged players often use a “guerilla warfare” strategy which stochastically allocates zero forces to a subset of battlefields. The advantaged player often employs a “stochastic complete coverage” strategy, allocating random, but positive, force levels across the battlefields. In the lottery treatment, where winning a battlefield is probabilistic, both players divide their forces equally across all battlefields. Due to the constant-sum nature of the game, we examine behavior under both random pairing and fixed pairing protocols. Under the random pairing protocol, players have significant serial correlation in allocations to a given battlefield across time. Under fixed pairing this correlation is significantly reduced, and disappears for the disadvantaged player in the auction treatment. JEL Classifications: C72, C91, D72, D74