Author: Howard Exner
Date: 21:49:18 01/05/99
Go up one level in this thread
On January 05, 1999 at 22:24:38, Enrique Irazoqui wrote: >On January 05, 1999 at 20:55:11, Matthew Herman wrote: > >>The position 8/7k/3R4/7P/6K1/4b3/8/8 w - - 0 1 between Tiger and Rebel10 leads >>to a won position in 11 moves as reported by the ken thompson egtb. >> >>Tiger might or might not find it but there is a definite win. > >Of course it's a win, but the question is that Tiger won't find it. Take a look >at the moves before and you'll see that it doesn't know how to proceed. After >the tournament is finished I'll replay the game from this position. Here is some analysis from an endgame book if others would like to test their programs. 1. Rg6 Bh6 (now the position reached is found in the book "Rook vs Minor Piece Endings", by Yuri Averbakh) The game is between B.Guretzky-Cornitz way back in 1863! The moves continue ... 2. Kf5 Bd2 3. h6! Be3 (if Bxh6 then Kf6 ... Kf7 sets up a win by the double threat of mate and winning the bishop. This, the only winning technique, could prove challenging for non tablebase programs as you have already mentioned). 4. Rg7+ Kxh6 5. Rg6+ Kh7 ( if Kh5 then Rg3) 6. Kf6 Bd4+ 7. Kf7 Ba7 8. Ra6 and so on. In playing over the your tournamant games I'm noticing more and more that the games are lasting longer between equally matched software. If one program would break away from the pack in regards to more advanced endgame technique then I'll wager that program would have a hefty lead in something like the SSDF list.
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.