The Search for GTO: Determining Optimal Poker Strategy Using Linear Programming

This project applies techniques from game theory and linear programming to find the optimal strategies of two variants of poker. A set of optimal poker strategies describe a Nash equilibrium, where no player can improve their outcome by changing their own strategy, given the strategies of their opponent(s). We first consider Kuhn Poker as a simple application of our methodology. We then turn our attention to 2-7 Draw Poker, a modern variant onto which little previous research is focused. However, the techniques that we use are incapable of solving large, full-scale variants of poker such as 2-7 Draw. Therefore, we utilize several abstractions techniques to render a computationally-feasible LP that retains the underlying spirit of the game. We use the Gambit software package to build and solve LPs whose solutions are the optimal strategies for each game