Author: KarinsDad
Date: 22:24:33 02/27/99
Go up one level in this thread
On February 27, 1999 at 21:45:58, Michael Ginat wrote: >On February 26, 1999 at 15:33:43, Dann Corbit wrote: > >>Personally, I don't see what Fischer and Kasparov have to do with computer >>chess. Quite frankly, I think Fischer is an embarassment if all the vicious >>things he is supposed to have said are true. >> >>If we had personal killfiles (hint, hint) every thread with 'Fischer' in the >>title would be in my own little toilet. I'm probably over-reacting since I got >>so horrified by some of the goings on in rec.games.chess.misc and round-abouts. >> >>In the words of Mr. Horse, "Well sir, I don't like it." > >I also am very sad that Fischer has said those things (especially as a Jew who >lost many relatives in the Holocaust). >On the other hand I think a discussion is relevant because:- >a) Kasparov makes extensive use of computers >b) Fischer did not and got to the same level or higher - a discussion could be >created on how computers are changing modern chess, the way we prepare and is >there evidence that people are stronger now with all that help? >c) Fischer proposed shuffle chess and that leads to the question of how strong >computers are without huge opening books. If I'm not mistaken the big guns like >Hiarcs and Fritz depend a lot on openings in deciding who will win when they >play eachother. Hi Michael, Computer programs are not that much different than humans. They both rely (for the most part) on good opening books. In the case of computers however, I think that there are two reasons why they are so dependent on opening books. One is that the standard tactics discovered by a good search engine do not correspond well to the opening (with some exceptions). This means that it is easier to use a small (several million position) opening book than it is to write good code to figure it out on the fly. The other, and more important reason, is that people have spent decades analyzing the openings down 15 or 20 moves (per side). This is a lot deeper than the search engines can look. This means that it is extremely unlikely that a computer will often discover many new things within the opening (not impossible, just unlikely). However, there are too many middlegame positions for humans to have analyzed (or remember) even a fraction of them, so computer tactics can prevail there (since the program and the human are often looking at a relatively "new" position there). From my limited experience, programs are not that powerful when it comes to the opening. However, the reason for that is that it is more cost effective to use an opening book than it is to come up with good algorithms to handle it. KarinsDad :) PS. When is this year's Colorado Closed? Did you play it already? >As for the goings on in rgcm I agree, but then this club bans the lunatics who >spoil the discussions with personal attacks, and I'm not worried that it would >degenerate to that level in this club. > >regards, >Michael
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.