Author: Robert Hyatt
Date: 08:59:27 02/26/01
Go up one level in this thread
On February 25, 2001 at 11:23:17, Enrique Irazoqui wrote: >Weeks ago I posted this problem of Troitzky. Why is it unsolvable? This is the >original position: > >[D]8/3kP3/3br1P1/6P1/7K/8/8/7B w > >and this the final position after 1.Bc6+! Kxe7 2.Bd5 Rxg6 3.Kh5 Rg7 4.g6 Kf6 >5.Bf7 Kf5 6.Kh4 Kf4 7.Kh3! Be7 8.Kg2 Bh4 9.Kf1!= =. Not deep at all. > >[D]8/5Br1/6P1/8/5k1b/8/8/5K2 b > >Here Crafty and all the others evaluate it as near +6. Since programs know that >KB vs.K is a draw, is it not possible to make abstraction of the trapped piece >and the white bishop and pawn that trap it, and then realize it is a draw? Step >1, identify the hopeless trap; step 2, evaluate "as if" black's rook and white's >pawn and bishop didn't exist. If such a simple case can't be seen by programs, >how can they avoid giving wrong evals when the theme of trapped pieces appear, >as in the Sherbakov-Rebel game and several others? It's probably a foolish idea >of a programming illiterate, but these cases are quite maddening. > >Enrique I agree, although this particular case is _very_ hard to handle for me. IE I don't really want to look all over the board to recognize such a trapped piece, I tend to reward the mobility of a piece as a way of avoiding such positions most of the time..
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.