Author: Robert Hyatt
Date: 14:39:18 12/10/03
Go up one level in this thread
On December 10, 2003 at 16:07:17, Gerd Isenberg wrote: >On December 10, 2003 at 14:55:20, Robert Hyatt wrote: > >>On December 10, 2003 at 12:05:43, William H Rogers wrote: >> >>>I have read and reread the rules of chess reguarding the three move repretition >>>and how it applys to a draw. Now maybe I am wrong but it appears to state that >>>if a person discovers a three move repretition that he can ask for a draw and if >>>its acceptable by both parties then it is to. No where does it say that it is an >>>automatic draw. >>>Let me illustrate the follow in this hypotical game. >>> >>>Player 'A' is a superior chess program and is way ahead on points and is assured >>>a win. >>> >>>Player 'B' finds that he can make a move that will give him a three move play >>>and asks for a draw. By doing so, he greately improves his overall point >>>standing by drawing against a superior program. >>> >>>However Player 'A' knows that if he accepts the draw he will lose points and he >>>should not as he has the 'game' escentially won, so he refuses and takes the >>>game to its logical conclusion. >>> >>>Draws were design to offer players of somewhat equal strength a way to end a >>>somewhat pointless game that can not necessarily be won by either side. >>>There are board positions in which according to the rules neither side can ever >>>win so draws are the only option left. >>> >>>Over the years I have noticed some people have assumed that because a rule can >>>apply in some positions that it should apply in all positions. At that I say >>>just read the rules until you completely understand. >>> >>>Bill >> >> >>There is no "agreement" required. If a position is repeated for the third >>time and it is my move, I may claim it instantly, and call the TD for >>verification. If the position is repeated for the third time _after_ I make >>my move, then I write the move down, but do not make it, and call the TD over >>for verification. My opponent has no input. >> >>If you want your chess engine to look _really_ stupid, then let it play >>through a draw by repetition and lose. Or just play an infinite game, >>hoping your opponent will screw up. >> >>You miss one important detail, since this is related to _computers_ only. >>You have _already_ repeated the position once and your opponent did not >>blunder. Do you _really_ think that if you repeat again and don't claim the >>draw, that he will blunder this time around? If so, how many cycles before >>you are convinced that he won't? >> >>That's why it is a 3-fold repetition and not a 2-fold repetition. Once lets >>you get back to an old position and try a new move. Twice says "you had your >>chance." > >Good point. I don't have any problem with an explicit draw rule for computer >chess - with or without operators, declared by the ICGA. To terminate the game >immediately, if a 3-fold repetition occures. I don't think that should be done. I believe that one of the two players should _correctly_ claim the draw according to the normal FIDE rules for doing so. Then the arbiter, in this case the FICS, can verify the claim and end the game if it is valid. Notice I said "one of the two players". I don't care if the claim comes from special purpose hardware, from the GUI, from the engine, from some daemon looking over the shoulder of the engine, etc. The "player" is the "computer system" simply and clearly. > >OTOH, if the rules don't change here for ICGA-events, and we have to play >according to FIDE rules, you must consequently leave it to the programs to claim >draws or not, and probably to play some funny games after gaining some more time >after move 40 or 60. There is still the 50-move rule ;-) I have no problem with doing that. But at present, with non-automatic interfaces, it presents problems. Programs don't do it right. It is hard to figure out which came first, the draw claim or the move, etc. Eliminate it by using a server that does it right. ICC is supposedly fixing this now so that to claim a draw after your move you send "draw ..move.." which matches the FIDE rules correctly. > >Another point to fix by ICGA rules is sharing of chess knowledge via commercial >or none open source GUICKs. That should be explicitly prohibited for none native >and none commercial engines. > Why they refuse to address this is a mystery. I suspect it is to avoid getting a commercial entry/set of entries mad... Which is not a valid reason, IMHO. >Gerd > >> >>In the ongoing discussion here, the computer said "this is a 3-fold repetition". >> >>That's good enough for any reasonable TD...
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.