Author: Robert Hyatt
Date: 13:34:22 11/11/98
Go up one level in this thread
On November 11, 1998 at 16:00:38, Howard Exner wrote: >Typo on the header. > >3r2k1/p4ppp/2p5/Np6/7N/BP1n2PP/P1b3B1/6K1 b - - id id "Markowski - Shirov"; > bm b4; > >Here is the other position from the Chess Life Shirov article. His comments are: > >"Again the same story. Playing against the computer, White could be confident >that 23 ... b4 would not be played." this is all a waste of time to read and comment on. Someone took several of these "positions" and tried them on Crafty... the results were posted on r.g.c.c a few weeks ago... a couple it got quickly, none took a day... He is grossly misinformed here... > >and later he states, "Maybe after some 10 years the program will find 23 ...b4" > >Well in this position Rebel in Default mode with Combination Off (as is Rebel's >setting when playing games) it took 17:29 to play b4. True this is not within >normal tournament time controls but who knows what would happen in a 40/2 >game. If Rebel was thinking on the opponents move and the opponent took >10 minutes to move with a correct prediction by Rebel as to the opponents move >then maybe it would be played. Also telling is how Rebel 10 came about playing >b4. Up to the 12th ply Rebel likes the move c5 ( ply 12.00 at 5:46) during this >time it does not pay much attention to the move b4 as it is 12.13 in the >iteration. By the time it starts to think of b4 it is 16:38 on the clock and >then shortly after it chooses b4 as best (17:29) and holds on to this move. >Machine is a K6-233 with 60 MB hash in Dos. > >So if b4 ended up higher on the iteration list it looks as if Rebel >quite possibly could have snared this one also in a tournament game. >Again, I'll wager that more than a few programs will find b4 in under 5 minutes. without a doubt...
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.