A Methodology for Learning Players' Styles from Game Records

TL;DR

This paper presents a methodology for learning individual chess players' styles from game records using temporal differences, with promising results in distinguishing players and potential applications to other strategic domains.

Contribution

It introduces a novel approach to infer players' styles from game data using temporal difference learning within a chess engine framework.

Findings

Successfully learned styles of two chess world champions

Able to discriminate players based on learned styles

Discussed limitations and future research directions

Abstract

We describe a preliminary investigation into learning a Chess player's style from game records. The method is based on attempting to learn features of a player's individual evaluation function using the method of temporal differences, with the aid of a conventional Chess engine architecture. Some encouraging results were obtained in learning the styles of two recent Chess world champions, and we report on our attempt to use the learnt styles to discriminate between the players from game records by trying to detect who was playing white and who was playing black. We also discuss some limitations of our approach and propose possible directions for future research. The method we have presented may also be applicable to other strategic games, and may even be generalisable to other domains where sequences of agents' actions are recorded.