Abstract
As part of an investigation of not necessarily optimal strategies, the current article examines some human reactions to being confronted with samples of the KQKR endgame. The observed human error rate is studied and is shown to be quantifiable in several dimensions. It is found that in some cases human players progress only at a fairly small fraction of the game-theoretically optimal speed. The observed behavior of human fallible players induces at least the basis for investigating non-optimal play by an agent wishing to entrap the opponent into a draw by the 50-move rule, the opponent having a win when playing optimally.
Get full access to this article
View all access options for this article.