Learning to Player Poker using Counterfactual Regret Minimization
In our investigation into Counterfactual Regret Minimization (CFRM), we attempt to determine Nash equilibria for Kuhn poker and Leduc poker. Both of these games are relatively simple examples of sequential, imperfect information, zero-sum games. Thus, they provide a good test bed for assessing CFRM’s ability to determine optimal bluffing strategies in similar, more complicated games like No-Limit Texas Hold’em.