An Empirical Study of Artificial Participants

Abstract

This study is an investigation of artificial participants operating in their appropriate environment. The artificial participants in this study are artificial chess players and their appropriate environment is a chess game. This study is an empirical investigation testing the hypothesis that the length of a chess game is inversely related to the difference in the chess skill levels of the artificial chess players. Five series of chess games of 18 games in each series were instituted between five pairings of web-based chess engines. The chess engine, chess.com Level 10, was a player in each series and won all 18 games in each series. The opposing players came from the Play Magnus app at five different levels of chess skill. This study provided an investigation of 90 chess games involving artificial chess players. The hypothesis for the study was confirmed. Game length was significantly inversely related to the disparity in chess proficiency between artificial chess players. This is one of the first scientific studies of artificial participants.