Author: Robert Hyatt
Date: 08:11:11 03/11/01
Go up one level in this thread
On March 11, 2001 at 10:44:14, Robert Hyatt wrote: >On March 11, 2001 at 02:25:42, Ferdinand S. Mosca wrote: > >>Hi Everyone, >> >>One of the games from the test between Deep Fritz and Crafty 18.01 >>using P3_667mhz_128mb with 32mb each and 3+4+5 EGTB at G10. >> >>An illustrative game to examine showing the strenght of Crafty >>handling a passed pawn from early middle-game till the end-game until it reaches >>the following position. Crafty played bxa6?. >> >>Deep Fritz - Crafty 18.01 (1/2-1/2) >>[D]8/1k6/p7/KP1p4/P2B4/2P5/n7/8 w - - 0 1 >> >>How many engines have progressed on this area? >> >>Tigers and Nimzo 8 seemed to take the rook_pawn. >>Deep Fritz, Junior 6, Hiarcs 7.32 and Yace does not take the rook_pawn. >> > >Crafty has code to prevent this. But apparently, somewhere during the past >few months, I broke it. It was definitely modified sometime back, and it looks >like I introduced a horrible hole in the code. After black takes the c pawn, >there is no way to win. And it used to know this. I am looking now. 18.5 will >restore this to working. > > Found the problem. I make sure that the "losing" king is closer to the queening square than the opponent's king. But I factored in wtm, and if my king is one square away from a8 in this position, and your king is 2 squares away but you are on move, it didn't handle it correctly. :) It thought your king could get just as close as mine, illegally. Bob > >>How do other programs evaluate this position? >> >>Regards, >>Dinan
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.