Author: Fernando Villegas
Date: 18:28:47 05/20/99
Go up one level in this thread
On May 20, 1999 at 18:59:56, William H Rogers wrote: >There is another exception. A few months ago, a game was won before the computer >ever left its book. Well, that is, when it finally got out of its opening book, >it ended up in its end game function and the game was won!!! >With the current progress in chess programs, I'll bet that most of chess will be >solved in a decade or two, at least serious chess. If we disreguard the useless >positons, then the normal game should be solveable. This is the key concept: "normal" game. It is related with the idea that chess should be considered not a sucesion of moves, but of positions. Of course, positions are reached with moves, but many moves of thread of moves give birth to the same position. In fact, if the atomistic view of the game -and so the idea of it as something infinite- was correct, then we could apply the same reasonning to any kind of physical phenomena. We could say that we cannot solve the equation of the fall of a body because there are some many atoms in different arrays into it that are different to any other body. Science and calculation are macroscopic and that's the reason they can give us laws of behaviour. Everything behaves differently and is unpredictable if we see it according a very microscopis angle, but that is not relevant for our purposes. Same with chess. So we can "solve" the chess game if we consider that solution does not means to fill an infinite hash table with all possible moves, BUT to grasp the esential law that govern the development of positions and win the game from that. In other words, if a device can win all the games or at least get a draw, then ches is solved no matter if you have in the pocket all the moves or you have the tool for a better handling of positions. fernando I think. >Then a little later, some chess problems were posted and the machines were >allowed to think > infinity... and many found mate in 30 or so. >As the compition in chess programming developes, chess is getting better all the >time. Don't forget that when the first calculations on the probability of >solving chess, the worlds biggest computer "ENIAC" only had 4k of memory. >IHM >Bill
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.