CHESS EVALUATIONS

In 2022, heated controversy stirred the world of professional chess: Magnus Carlsen, arguably the greatest player of all time, dropped out of a major tournament after losing in an upset to Hans Niemann, a young up-and-comer at the time. He insinuated cheating, of which, at least in online games, Hans Niemann had had a history, and the chess world subsequently went haywire. In their next match, this one online, Carlsen resigned after making a single move. More fuel, more flames. Pundits all over were commenting, some in support of Niemann, some in support of Carlsen. Analyses of mixed quality made their rounds, some apparently confirming, others dismissing the cheating allegations. The most viral analysis was carried out by a Twitter user known as "gambit-man", and linked and commented on prolifically. It alleges that, using Engine-Game Correlation, that is, the percentage of moves made that correspond to the best engine move in that situation, one could nearly conclusively show that Hans Niemann did, in fact, cheat. Not one to shy away from trying to replicate an interesting analysis, I took a look of my own. Programmatically, reproducibly, in Python.