Author: Robert Hyatt
Date: 09:54:08 12/01/03
Go up one level in this thread
On December 01, 2003 at 12:36:35, Francesco Di Tolla wrote: >On December 01, 2003 at 10:28:23, Robert Hyatt wrote: > >>On December 01, 2003 at 04:18:00, Francesco Di Tolla wrote: >> >>>As far as I understand Jonny did not claim draw at all. >>>Not because the program did not ask for it while the interface did, but becasue >>>both the prorgam and interface NEVER did a draw claim. >>> >>>The interface showed a pop-up claiming "3 fold repetition" and this is NOT a >>>draw claim. >> >>Then what is a "draw claim"??? > >That should be agreed, but something like "the engine asks for a draw due to >..." or the like would be enough. You don't "ask for a draw" in this case. It _is_ a draw. It simply has to be validated by the TD if the opponent doesn't agree. Both may agree even before the 3-fold repetition if they choose, in human play. > >"Info: 3-fold repetition" does not express the intention of the program. > The program has no "intention". It simply analyzes and "makes a statement" That is absolute truth. >> >>I have been a TD for 30+ years. I've run hundreds of tournaments, with >>and without computer players. I don't have any specific wording that is >>required to claim a draw. > >I have participated to international tournaments an I can assure you that if I >wanted to ask for a draw I had to explicitly do it with the referee. I have been to US Open events (more than one). Players agree to draws all the time without a TD's involvement in any way. I say "draw?" You nod "yes". I offer you my hand. We shake, sign each other's score sheet with the draw result, and we are done. The TD has no part in that. > >If I call the refere and say "3-fold repetition" he can respond "So what you >mean". > And I can respond "3-fold repetition". I don't have to know his language even. >>Nor do I find any such wording requirement in >>my FIDE rule book. I have had a player look at me, hold up 3 fingers, and >>motion me over. _That_ was a perfectly valid 3-fold rep draw claim. I looked >>at the score sheed, and said "yes". > >No that was a draw by agreement due to the fact you knew he could go to the >referee and ask for the application of the rule. If he did not this is just an >agreement of the two of you. Of course this is a courtesy I would apply as well, >but I would not execute my third move without the refree there and having You miss the point. _I_ was the TD. I was in the playing hall, simply walking around and looking. The player held up 3 fingers and motioned me over. I _knew_ what he wanted. >expressed to him my intention clearly, cause my opponent could be less >"gentleman" then you are and after the move was executed pretend to play on. >This would not be the first nor the last time it happens. > >>The computer (the monitor, the cpu, the memory, the gui, the engine, the >>opening book, the endgame tables, the disk drives, the keyboard, the mouse, >>and so forth) said "this is a 3-fold repetition". Do you _really_ think it >>better to say "I claim this is a 3-fold repetition". Or "I want to claim >>a draw by the 3-fold repetition rule" or whatever? > >Sure that is what I said. I mean, I'm making a paradox of this issue. As a >Shredder operator I would not have pretended to continue claiming my opponent >did not formulate the request properly, since as you noticed almost no program >odes it. > >But this should be fixed: thre should be a clear consesus oon how theprograms >state the request. And somebody could complain that Fritz pop-up is not explicit >enoguh. > In computer chess, "3-fold repetition" is more than clear enough. The computer didn't just have an idle "senior moment" and spurt that out. >>"3-fold repetition" would be enough for _any_ sane person. In FIDE >>events there are _no_ specific words required. In fact, I might not even >>be able to speak to the arbiter directly, due to a language issue. I might >>need to ask an interpreter to inform the arbiter of my claim. > >3-fold repetiton is a generic statement about the position, as mate in 10 showed >in a line. >Should we argue that an engine noticing to be mated in 10 moves is resigning? No. But if an engine notices "mate" there is no need to say more, is there? By the rules, mate ends the game. As does 3-fold repetition when a player points it out. Both _can_ play on if they choose, but if one says "this is a 3-fold repetition" he can end the game then and there. Mate in 10 does not end the game. Mate does for sure. repetition or 50-move rule draws end the game if a player wants. > >What if the an engine would write a generice 3fold repetition in its evaluation >line and continue it's analysis and play on. How should we interpret this? > the analysis means nothing. What counts is what the engine says "on the board". It popped up a window saying "this is a 3-fold repetition". It must have meant to do so. Just because it shows up in the analysis is meaningless, because it might find a better move that avoids the repetition. >>Correct. The computer "player" said "this is a draw". >>That's enough. > >no the word draw was not mentioned at all. "This is a 3-fold repetition" == "this is a draw" by the rules of chess. "mate" == "win" by the same rules. > >>>Actually if we strictly follow FIDE rules the operator can ask a draw for 3-fold >>>repetition only if the engine does the explicit claim >> >>That is wrong, there is no such FIDE rule at all. > >there is it is down there No. The FIDE rule says "only if the computer offers the draw". FIDE (nor any other sane person) tries to distinguish between the GUI and the chess-playing code. In some programs there is no distinction. IE when I play in a tournament, I use console mode where there _is_ no GUI. Bruce (Ferret) has a built-in GUI that is a part of the Ferret source code. > >> Nor is there any FIDE rule that explicitly gives the wording >> required to claim a 3-fold rep draw. > >that's to be clarified I agree > >>>"3.3 Only if the computer itself so instructs him may the operator offer a draw, >>>or claim a draw by repetition." >> >>The "computer" did this. _clearly_. > >this is not clear to me. Please understand me, I wanted to underline the paradox >that the whole issue is unclear to me. > >>What is a bug here? The gui says "this is a 3-fold repetition". It couldn't >>be any clearer or any more correct there. > >to me it is not clear: is it asking or not to apply the rule, or stating >something or what? It is stating a fact. "this is a 3-fold repetition. I noticed it. The game is there a draw." I suppose it could quote the entire FIDE rule pertaining to this. But that is not exactly how I would program a computer to annouce a repetition. Crafty says this: Print(128,"%sgame is a draw by repetition.%s\n" Print(128,"%sgame is a draw by the 50 move rule.%s\n" Print(128,"%sgame is a draw due to insufficient material.%s\n" I don't think the "game is a draw" part makes this any more (or less) valid than if I had just said "this is a three-fold repetition" There can only be _one_ interpretation. As I said, computers don't just randomly puke up some sensible-sounding string of words, for no reason. > >>The operator can't say _anything_. That is the point. The operator is _not_ >>part of the playing system. > >the opoerator must say something: he must go to the referee and say: my engine >asks for the application of.... but this was not the case. > Correct. The operator made an error. Look at the ICGA rules for what happens when the operator makes a mistake. "The game is restored to the point of the error and the error and clocks are corrected." The game continues from that point. Or in this case, the game ends because correcting the error ends the game with a draw result. >>>Who could claim the oppposite? >> >>Any rational person? > >I'm not stating I would do that, I claim some could ask to follow a rule to such >limit, that's the paradox I wanted to stress, I understand your "common sense" >and agree with it, but sometimes clear rules prevent such misunderstandings. Yes, but sometimes a TD can take a clear rule and turn it into muddy water, in a heartbeat. :)
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.