Analysis of Trends in Chess Games Between Selected Grandmasters

Chess is a deep and well-studied game with a wealth of content and data. Because of the relative simplistic nature of the game states, games themselves are relatively easy to record, and thus there are large databases of games, dating back to Francesco di Castellvi vs Narciso Vinyoles, 1475. The vast majority of published games are between masters and grandmasters, usually recorded by tournaments; recorded games of individual players are also often collected throughout their lifetime and later compiled. In this paper, we aim to perform basic analysis of several trends over time: opening moves and their replies, the win rates of knight-bishop imbalances, and the win rates of uncastled sides. and create simple models of behavior based on the era of play. While this is by no means a complete analysis, our goal is to demonstrate several techniques of statistical and data analysis applied to chess and quantitatively show some trends which were previously assumed to exist by the community.