Author: Dan Benjamin
Date: 17:10:49 04/12/03
It is, of course,unknown whether chess is ultimately a win, loss, or draw for white, and wecannot know the answer for sure until chess is solved. However, it may be possible to gather some suggestive evidence: The idea would be to have a computer solve different varieties of mini-chess on a 4x5 chessboard, 5x5, and, if possible, 5x6. For the larger boards, it might take a computer running several days to solve it, but it might be doable. In each case, the set-up of white and black pieces would be symmetric (e.g., NQKB in the 4x5 case), but one could vary the set-up (e.g., trying BQKN, RQKN, etc.). The questions would be: Which of these are wins? Do the winning possibilities fall with the size of the board? Does black ever win? Of course, it may be difficult to infer anything about 8x8 chess from mini-chess. On the other hand, certain possible outcomes of this experiment would be strongly suggestive. For example, suppose it was found that all versions are draws. There is little reason to suppose that the outcome would be different in 8x8 chess. In any case, the games themselves might be entertaining. And it might be fun to resuscitate a research program on mini-chess, which was used extensively in the early days of computer chess. If anyone wants to run these experiments, I would be very curious about the outcomes.
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.