Author: Uri Blass
Date: 01:20:25 12/01/00
Go up one level in this thread
On November 30, 2000 at 18:36:23, Mike S. wrote: >On November 30, 2000 at 05:24:15, Uri Blass wrote: > >>(...) >>>>An elegant computer move: >>>>[D]3k4/p7/K3BP2/8/7p/8/2P4P/8 w - - bm Kb7 >>>>1.Kb7!! © WChess 1.04 >> >>I believe that there is an error in this position because Kb7 is not the only >>winning move. >> >>When I look at the diagram I cannot find a move for white that is not winning. > >Of course you're right that white wins anyway, but Kb7 s the fastest way to the >mate. The position is from a computer game WChess 1.04 (485/50) vs. Excel >MachIIc. Simple chess programs, but many human players too, would just play Kxa7 >I think. At least for some time on the beginning of the analysis. As you write >in the other posting, it is the idea of keeping the bK in a cage (instead of >just capturing a pawn), which makes the move look "humanlike". > >It is also not as easy as it may seem, although only 11 ply deep. On PII/333, >the following programs took more than 1 minute to find 1.Kb7: > >LiGo2000a, CSTal (DOS), Junior 5, AnMon 5.01, Bringer 1.5 (1.7 finds it), Comet >B11 (B02 finds it). > >But the majority of progs find it fast. I believe that part of the programs that cannot find it in 1 minute will never find it because they find another move that leads to mate because of extensions and after finidng one move that leads to mate they are not looking for the fastest mate. I do not think that the position is a good test position because not finding the fastest mate is not a serious mistake because you do not lose even 0.00000000000000000000000000000000001 rating points because of this mistake. 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.