Author: Miguel A. Ballicora
Date: 22:04:10 05/10/01
Go up one level in this thread
On May 10, 2001 at 21:57:14, Dann Corbit wrote: >On May 10, 2001 at 21:41:31, Robert Hyatt wrote: > >>On May 10, 2001 at 20:15:19, Dann Corbit wrote: >> >>>Consider this position: >>>[D]r1b1r1k1/p4pp1/2p4p/n7/2PP4/2N2Nq1/PPQKBnP1/R1B3R1 b - - >>>Black is behind in material. Black has a mate in 4. >>> >>>Suppose at some juncture of the game black could claim checkmate by the 50 move >>>rule if black so desired. Would your search algorithm claim the draw or ignore >>>it and go for the checkmate? >> >> >>Crafty searches it as a zero-sum game. If it sees a forced mate, it realizes >>that the opponent will claim the draw at move 50 if he can... Crafty would not >>walk into the draw voluntarily unless it could find nothing better (ie it might >>try a pawn push or capture that resets the 50-move counter without giving up >>the forced mate opportunity. > >I mean it the other way around. The opponent has done a 3-repeat or the >opponent has gone 50 non-reversible moves. It's your turn to play. Will your >program seek the mate or claim the draw or something else My program would take the draw. Regarding the 3 rep, it is ok because it will never be in that situation in the first place. If the opponent repeated 3 times the position in the move just played, it is because my program allowed it. In other words, my program would have checkmated already. However, regarding the 50 moves repetition, it is a flaw taking the draw immediately. I haven't attacked the problem because there are tons of things in my to-do list that require more attention. It is a rare case, but an elite program should take care of that. Mine is not elite :-) Miguel
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.