In 1950, Claude Shannon published his seminal work on how to program a computer to play chess. Since then, developing game-playing programs that can compete with (and even exceed) the abilities of the human world champions has been a long-sought-after goal of the AI research community. In Shannon's time, it would have seemed unlikely that only a scant 50 years would be needed to develop programs that play world-class backgammon, checkers, chess, Othello, and Scrabble. These remarkable achievements are the result of a better understanding of the problems being solved, major algorithmic insights, and tremendous advances in hardware technology. Computer games research is one of the important success stories of AI.
Samuel's successes included a victory by his program over a master-level player. In fact, the opponent was not a master, and Samuel himself had no illusions about his program's strength. This single event, a milestone in AI, was magnified out of proportion by the media and helped to create the impression that checkers was a solved game. Nevertheless, his work stands as a major achievement in machine learning and AI. Since 1950, the checkers world has been dominated by Tinsley.
Polaris, a rising star in the poker world, has professional card players fretting. The 16-year-old has a perfect poker face, can shift strategies in an instant, and never gets tired. Well, you do have to recharge the laptop batteries from time to time. This crafty computer program is one in a long line of codes designed to compete with humans. And for many games, machines now surpass even the best human opponents.
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.