The Role of Computer Programs in Checkers - A Programmer's Perspective

Early attempts to program machines to play checkers were hindered by the slow hardware available, but as computers have become faster and search algorithms have improved, the standard of machine play has improved dramatically. Now anybody with a PC can have access to an indefatigable electronic checker grandmaster at any time. This has been seen as both a curse and a blessing by human checker players. There is no doubt that, for the serious student wishing to improve his or her game, a strong sparring partner is invaluable. However, the widespread availability of checker software, especially freeware or shareware programs, is open to abuse in real-time internet checkers, where unscrupulous people can pretend to be stronger players than they really are by using a checker program to choose their moves. In Mail-Play the use of computers is allowed, even expected, and has no doubt reduced the frequency of tactical oversights and thereby elevated the standard of play even higher than it already was. Nevertheless, mail players know all too well that the every program is fallable. It is the combination of human skill and computer calculation that is the recipe for success here. In over-the-board play computer programs tend to be excluded from competition on the grounds that they are too strong and people do not want to play them. In the field of analysis computer programs have played a major role in discovering mistakes in published play and finding new resources in supposedly untenable lines of play. This can surely only be for the good of our Grand Old Game.

I hope that in creating Wyllie I have contributed something good to the game. In order to continually develop and improve a checkers program it is necessary to let the program play against strong opposition, just as human players need difficult oponents if they are to improve. In the process of development and fine-tuning Wyllie has played thousands of games against human oponents. Many of these players have also learned new things about the game, demonstrating that human versus machine contests can be beneficial to both the human player and the program developer alike. I wish that there were more opportunities for program authors to test their creations in a competitive arena, and I commend Vinco Online Games (VOG) for their regular "Player v Machine" tournaments which are open to human players and programs entered by their authors. As a matter of interest the record shows that the humans actually are a match for the machines, as many of these events have been won by human players in spite of the silicon threat.

In face to face events it is a different story. Players have paid to enter a tournament and may have travelled long distances. Organisers of such events usually do not allow computer entrants as this would perhaps deter much needed human players from entering. However, even here I feel that there is scope for a mutually beneficial interaction between programmers and human players. For example, what if there is an odd number of human entrants? This means that in every round someone has to receive a "bye", that is they receive a full point as if they had won, but they do not get to play any draughts. A computer oponent could at least provide a game for the player with the bye, and the player could still be awarded his point regardless of the outcome.

In the game of chess there are occasional "Man versus Machine" tournaments and matches. I feel that there is a place for such events in draughts as well. Personally I suspect that in a tournament consisting of the top commercial draughts programs and the top human master players the humans would do rather well, especially if they prepared "anti-computer" lines of play in advance, a well known strategem for human chess masters. The resulting play would undoubtedly be of a high standard and would surely enrich the literature of the game. The novelty of such an encounter might well attract much needed publicity for the game and thereby also for the governing organisations such as the ACF and the EDA. I hope that the ordinary rank and file players can look upon programmers such as myself as fellow enthusiasts for this great game rather than just a competitive menace.

Send E-mail If you have any opinions about any of the above I would be interested to hear them - please feel free to email me.

Roberto Waldteufel