Author: Robert Hyatt
Date: 10:10:20 07/29/98
Go up one level in this thread
On July 29, 1998 at 07:54:07, Guido Schimmels wrote: >As FIDE rules were not designed with computers as possible opponents in mind, > they don't apply to computers. In fact computers break the FIDE rules in many > ways, if you think of it: >1) They don't note the moves on the form >2) They don't move the pieces nor handle the clock >3) They access external information during the game (opening book, tablebases) > (3) is an old argument not really worth discussion. What do you do with a *human* that has perfect memory? IE I have personally watched several different GM players go over a game they played 10 years ago, move by move, sub-variation by sub-variation, without getting lost in the process. Browne is quite good. So is what me memorizes unusable? If so, I memorized hundreds of opening lines years ago. Are they unusable? If so, then why can't a computer do the same? It just has "better" memory... This is an argument that won't ever be ended... >Ok, 1) and 2) could be easily addressed, but what about 3) ? > >If I played and Open tournament and had to play a computer, >if I would use my opening and endgame library during the game and analyse on >a little extra chess-board how could the arbiter say I'm cheating if I only do >what >the computer does itself ? Actually, the computer is only doing what you yourself do. IE do you know what to play after I play 1. e4? Do you have an instant move from memory or do you have to think a few mins to work out the variations? > >I really think it is very important for the computer chess community to agree >on special rules for computer-human play - or what sense does it make to >discuss if computers are already GM level or not if we don't define the >underlying game conditions ? And the game conditions have a *huge* >influence on the rating we observe. >When Bob negates micro's to be on GM level, his typical argument is: >Look at ICC, computer's have still big weaknesses which players on ICC >will find out quickly and then humiliate them. >But are the conditions computer play on ICC fair or not ? I don't know, at least >they are very much different from for example the Kasparov vs. Deeper Blue >mach. Deeper Blue surely benefited highly from the human interference between >the games and from the special opening preparation against Kasparov (ok, >Kasparov >deviated from his usual opening play, but so the threat was stronger than the >execution !) There wasn't a lot of intervention between games. But I do that all the time on the servers. When I see something bad happening, I may well fix it before the game is over. I have been to tournaments where someone would say "would you help me look at this opening a minute? I know my opponent will open with 1. c4 and I'm not very familiar with that." Is *that* person cheating by being "modified" between rounds? > >I hope I could make my point clear. >So when we talk about computer strength, which conditions do we talk about ? > >- Guido - same as we talk about for humans, in fact. A computer and a human are basically different. But they have similarities as well, as in memory...
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.