Artificial Intelligence and Hold’em, Part 1: Counter-Factual Regret Minimization

It has been a good year for machine learning and artificial intelligence. And no, I’m not just referring to the Russian scientist who claims his Robotchef can make a perfect bowl of crab bisque.More relevant to the poker world, this was the year that researchers at the University of Alberta claimed to have “solved” heads-up limit hold’em, calculating a strategy for all possible cases of the game within 1% of an unexploitable Nash equilibrium. Meanwhile the poker group at Carnegie Mellon University in Pittsburgh have tested a heads-up no limit hold’em system against some of online poker’s best heads-up players, a match that the humans won but only by a statistically inconclusive margin. These results have been covered here before. What I thought we’d do today is to dive into…


Link to Full Article: Artificial Intelligence and Hold’em, Part 1: Counter-Factual Regret Minimization