Author: Uri Blass
Date: 09:26:00 01/13/01
Go up one level in this thread
On January 13, 2001 at 11:52:00, James T. Walker wrote: >On January 13, 2001 at 08:10:47, Uri Blass wrote: > >>Fritz6 allows Steve Ham to repeat the position and I wonder if Fritz6 knows not >>to evaluate repetitions as a draw. >> >>You can find the game at >>http://correspondencechess.com/campbell/ham/fr_hambl.htm >> >>Steve Ham can repeat the position by Kc8 and get the exact position that was on >>the board after move 61. >> >>I have nothing against Fritz6's move(I think that the move is the only idea to >>make progress but it seems strange if Fritz6 knows it because usually programs >>do not know it and Deep Fritz is one of the programs that does not know it. >> >>Crafty knows it but Crafty is an exception. >> >>Uri > > >Hello Uri, >Because of this post I decided to look at the game. It is very interesting but >it brings up some questions. >Is Mr. Ham getting the PV/analysis from the operator? It seems according to his >comments that he is. If this is true then I'm wondering at what point in the >game is he receiving the Fritz analysis. He also seems to be receiving the >input of others who are following the game. Is he getting this input in time to >be useful? >As to your question I'm not sure how to find the answer. They have already >repeated a position at move 64 & 66. Repeating a position at move 71 equal to >move 61 would seem harmless also. I'm sure Fritz will not repeat either >position again unless there is no choice. This type of action seems normal for >programs in endings where they can't find a winning line. They seem to move >aimlessly around hopeing to "discover" something that did not show up in >previous searches. They usually do this untill the draw is forced or they >happen upon the right position that puts a win within their search. >I did see many moves by Fritz which were found after only seconds on my >computer. Of course that is not unusual either. In this ending I have seen >Fritz hit 34 plies in a few seconds with a score of 0.00 when looking at Blacks' >move. However when making the 0.00 move for Black Fritz then goes back to it's >2.xx score for itself. I believe Fritz "knows" the worst it can it is draw but >is programmed to continue looking and "hopeing" for a win. >Jim I find that my Deep Fritz evaluates the position as advantage for white and if I give it a lot of options it evaluates the move of the game as a draw. Re7+ with Ka6 is the only plan for white to make progress so I do not criticize fritz6's choice but I am surprised that it can let a repetition without evaluating the position as a draw(Crafty is the only program that I know that can play like this). Uri
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.