Author: Mogens Larsen
Date: 08:23:35 06/23/01
Go up one level in this thread
On June 23, 2001 at 10:15:53, Robert Hyatt wrote: >No he isn't. What do you do with a program that has a huge array with all >chess moves precomputed and just copied from the array when it is time to >generate moves for a specific piece on a specific square. IE Carl Ebeling's >thesis "all the right moves" was based on hardware that did just this. What >about hundreds of evaluation "patterns" that are stored directly in the >program and matched when an evaluation is done. Well, I forgot the hardware restrictions. Not on the number of CPUs and speed, but RAM and storage facilities. If the program you mention can function under the determined conditions then okay. That is, if the precomputed moves are made by that program. >And finally, what about humans that have memorized thousands or tens (or >hundreds) of thousands of moves and can recite them back perfectly? The likelyhood of a human player to have replayed the moves sometime on a chessboard is sufficient to allow human memory ;-). Besides, it's obvious that humans and programs don't play the exact same game. If it's supposed to hold interest, a balancing of strengths and weaknesses is important. >If crafty could simply use every game it has ever played, and I have most >of them, that would make a formidable book. But I don't see why that would >be any more acceptable than looking at what others have played, unless the >GMs are given the same limitation. The reason is that Crafty is handed the solution, ie. "Playing e5 now is the right move". Instead of having to figure it out by itself though play. You do that to a large extent, because your book is generated by pgn and not handtyped. However, Crafty is still guided by statistics of the most common move. If Dann is right about the 300 ELO, which I doubt, then it's a significant advantage. An advantage without effort. The same as endgame tables. That's nothing to be proud of IMHO. Regards, Mogens
This page took 0.01 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.