Author: Sune Fischer
Date: 04:39:37 12/11/03
Go up one level in this thread
On December 11, 2003 at 01:00:44, Terry McCracken wrote: >> >>_I_ did not attack anyone. So I don't know what you are talking about. > >You were not too friendly to say the least, and others were worse. Perhaps not, but you can expect replies in the same tone you use yourself. I won't quote his post again, but it was quite provocative and condecending IMO, not to mention wrong. > He >>presented a conclusion that was mired in inaccuracies and bad assumptions >>(He assumed FIDE rules were used when they are not.) His errors were pointed >>out by me and several others. > >Yes, most who don't know a tenth what he knows! BTW the FIDE Rules should be >enforced, and since they aren't in all situations, this should be changed. FIDE rules doesn't work for computer chess, that is one of the big problems. According to the FIDE rules drawing by 50 move rule or 3 fold repetition is an option legal by one of the players. Now suppose you have two programs which for some reason doesn't want to take the draw, what happens? I'll tell you. They are going to play 3000 moves until one of them crashes! - no I am not kidding, try it yourself! That's the technical side of it, it probably takes a chess programmer to fully appreciate these problems and realize their severity, no offence. But going a bit deeper than that, how do you design a program that knowns when to make the draw claim and when not to? I guess you want some sort of clause, "If I have a bug and accidentally played into a draw but otherwise is winning, then don't claim it". Sorry, but that is nonsense, it really is. Again, I think you need to get your hands dirty in programming and understand what alpha-beta can and can't do to make a correct ruling here. >If he made an error it was within the context of the ICGA rules. He's right >about the CC Olympiad. He's an expert at the highest level as a TD. >BTW did you check his credentials? He's no rookie, he's had plenty of >experience. > >He really does know what he can and can't do. > >He's a programmer like yourself, and has developed Poki, the worlds top Poker >Programme, and he is computer scientist with a PHD and to top it off a Canidate >Master, who has arbitrated many high level tournaments, including computer >chess. I can only respond to the post, how many degrees he has should only make his argumentation that much stronger. My impression was that he did not to know that much about computer chess. >If he had caught the error himself, he would have called Shredder vs Jonny a >draw, unless the ICGA said differently. > >But he pointed out that 3rd repeat does not have to be a draw, however in the >case of ICGA he normally wouldn't have the right to ignore the proplem or make >an official decision on his own to say play on. But since the game wasn't caught >in time he can rule with the body of the ICGA, if everyone is compliant to say >the win stands, which apperently he and they and all the rest were. >In cases where there is really big problems, he's right that as an arbitor >he does have a say, even in the ICGA if the ICGA allows him to exert his >authority. I disagree, for two reasons. First of all no FIDE TD would ever allow a player to lose on purpose, for that reason I think a draw should have claimed. Secondly, the draw claim by Jonny was valid. True his program did not know about the second repetition, but neither does any UCI program and perhaps some of the chessbase native engines are constructed in the same way. It is not the TDs job to disassemble the program, the 'solution' with the engine playing through the adapter in the chessbase GUI _did_ work, and the solution was approved before the tournament started. If Jonny wasn't allowed to claim a draw there then Jonny wasn't allowed to claim a draw at all. >However, this can be argued till the Cows Come Home. >> >>He should hardly expect to waltz in, make some comments that are really not >>relevant to what we are talking about, and waltz out without any counterpoints >>being made. > >He sure can, if he was asked to post the ICGA decision! And he did! > >As far as FIDE Rules are concerned, he can enforce them and if Computer Chess is >exempt from FIDE Rules, that should change as of now! > >It's absurd that a computer can't be subjected to the Official Rules of Chess! > >Third Repeat is the most tricky rule in chess and isn't mandatory, unless the >draw is claimed. For computers this hasn't been addressed properly within the >ICGA. > >Three fold repition of position is IMO the hardest rule to make clear, even more >so than en passant! > >It's of course optional too, and in this case computers haven't a problem in >this area. It _isn't_ optional for computers, if you make it optional you introduce a bug. Humans are sane creatures, they aren't going to play 3000 moves when it is obviously a draw. Computers only do what you tell them, and if it's not a rule to claim the draw then for sure sooner or later two engines won't claim a draw and the game will never *ever* finish. -S.
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.