First of all, the title is at least misleading. Heads-up limit hold’em (HULHE) poker, a two-person Texas hold’em game with limitations, is a small subset of the Texas hold’em universe. When the article was proposed for review, it seemed counterintuitive. An imperfect information game is solved? Amazing! Not so fast. In the article, the authors pull back from the claim of the title and provide three definitions of solvability, to wit ultra-weakly solved, weakly solved, and strongly solved. These definitions progress as one would expect, with the proviso that an imperfect information game resists a strongly solved solution. Thus, the title should instead be: “Heads-up limit hold’em poker is weakly solved.”

Nevertheless, this work demonstrates the authors’ heroic efforts to achieve their results--200 powerful computational nodes and 110565 subgames, each taking 61 minutes per game and run over 68.5 days--all to “obtain at least the game-theoretic value, for both players, under reasonable resources.”

Player anxiety, nervous tics (“tells”), and other human traits are not explicitly considered. Instead, a human competitor would face an impactable monolith with no skin in the game and almost complete insight into the various decision trees. No sane person would bet the family’s milk money against such an opponent.