Do Not Discretize, Optimize: Almost Greedy Fictitious Play
Provides a practical alternative to discretized Fictitious Play for machine learning applications requiring fast convergence in zero-sum games.
The paper introduces Almost Greedy Fictitious Play, a new variant of Fictitious Play for zero-sum games, achieving an instance-dependent convergence rate of O(1/T) in duality gap, matching Continuous Fictitious Play without discretization.
Our work revolves around Fictitious Play, one of the first iterative methods that is known to converge to a Nash equilibrium in zero-sum games. In recent years, there has been a revived interest, due to applications in various machine learning problems, which has motivated a line of work on its convergence properties and on proposing new variants of the initial algorithm. Our paper is along this direction and introduces one new variant, which we refer to as Almost Greedy Fictitious Play. The proposed algorithm greedily attempts to find the optimal stepsize at each iteration but its search space is constrained and includes almost all the line between the cumulative mixed strategy and the current best response. Our main result is that the method achieves an instance dependent convergence rate of $\mathcal{O}(1/T)$ with respect to the duality gap. This matches the rate of Continuous Fictitious Play, and offers an alternative to discretization. We complement our theoretical findings with experiments that demonstrate the effectiveness of the method.