Author: margolies,marc
Date: 19:02:28 04/29/04
Go up one level in this thread
Doctor Hyatt, your comments are fair. My own perception of the failure of my previous statement is that I allowed a 'subjectivity' argument to enter the mix. This happens because I did not define **progress** mathematically. The little mathematician inside me ( he should pay me rent!) tells me I should state that a definition of **progress** in an endgame is possible and leave it at that. But I'll go further to offer one with the expectation that the better minds here will shoot it down. We can use tablebases and computers to craft(no pun intended) a **progress** function. each position under question can scored for its distance to win lose or draw. If over a pre-determned large number of moves (say 50)the arbiter can simply demonstrate randomness of that number in a histogram then it's over. a draw. -marc On April 29, 2004 at 15:06:48, Robert Hyatt wrote: >On April 29, 2004 at 12:57:39, margolies,marc wrote: > >>Sune, I feel that your sample argument neglects the role of a responsible FIDE >>arbiter in a world championship match. Someone must see progress even if it is >>not involving the exchange of pieces of advancement of pawns. If progress can >>not be shown-- and tablebases will help here-- a draw claim ought to be upheald. >>It is in fact a sad state when an arbiter is unfamiliar with complex endgames >>yet must deal with claims of competing grandmasters under current time controls. >>this has happened in the usa quite a bit. I have heard stories-- but I will not >>repeat them and name individuals. > >Why does the arbiter have to understand the endgame in question? His function >is solely to enforce rules everyone has agreed to play by. The concept of >"progress" is not easily definable in some endings. Some humans have studied >some of the endgame tables and given up on understanding some simple plan to >reach the win... So how could an arbiter decide if progress has been made? > >Of course a computer could do this. But no human can play perfectly, and no one >would understand whether "progress" is real or not. IE it might be possible to >make progress until some key position is reached where it is very likely the >human will go wrong... > >subjectivity simply has no place in chess, IMHO > >> >> >>On April 29, 2004 at 11:42:49, Sune Fischer wrote: >> >>>On April 29, 2004 at 10:48:52, Helmut Conrady wrote: >>> >>>>On April 29, 2004 at 10:32:45, Vincent Diepeveen wrote: >>>> >>>>>On April 29, 2004 at 09:02:39, Victor Zakharov wrote: >>>>> >>>>>>>What do you think about the 50 move rule in computer chess, Guy? Any chances to >>>>>>>modify it for computer games? >>>>>>> >>>>>>>Helmut >>>>>> >>>>>>My opinion that rules will be fixed one day when tablebases will be more common. >>>>> >>>>>A 25 move rule would be more approriated indeed as the 50 move rule. >>>>>50 moves is just too long knowing that it just causes forfeits at the internet >>>>>online chess player. Other uses the 50 move rule doesn't really have. >>>>> >>>>>Shortly FIDE had extended it to 75 moves, but that was changed back soon >>>>>afterwards to 50 moves again. >>>>> >>>>>I doubt they will ever change it. >>>>> >>>>>If they change it, it will be 25 moves sooner than 75 moves. >>>> >>>>IMO that would contradict the spirit of chess. Why not allowing to win a won >>>>position? >>> >>>Chess is defined through the rules, if the rules say it's draw then it is draw >>>and not a won position. >>> >>>If you remove the 50 move rule you will see players torturing the opponent for >>>900 moves in a drawish RN-NN endgame. >>>They know it is a won position so they won't ever grant a draw, they just can't >>>figure out how to win it. >>>Imagine if this is an important game for the world championship, the game will >>>never end. >>> >>>That is not the spirit of chess IMO, you need a rules to prevent the game from >>>continuing forever. >>>But the solution is simple, if you have a winning advantage then don't trade >>>into an endgame you can't win in less than 50 moves. >>> >>>I for example try and avoid the KBN-K endgame because I'm not sure if I can do >>>it in less than 50 moves. >>>That is correct spirit of the game IMO, it wouldn't be any fun if I got a >>>trillion chances to win that endgame, he might as well resign on the spot. >>> >>>-S. >>>>Helmut
This page took 0.01 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.