Author: James Robertson
Date: 14:14:35 05/02/99
Go up one level in this thread
On May 02, 1999 at 06:32:47, Francis Monkman wrote: > >On May 01, 1999 at 23:30:58, James Robertson wrote: > >>On May 01, 1999 at 20:15:54, Francis Monkman wrote: >> >>>During a game with Genius4 in 95, the following position arose >>>after 17 moves by each side: >>> >>>White: Pa2,b2,c4,e5,f4,g2,h2 Ng3,g5 Bc1 Ra1,f1 Qe2 Kh1 >>>Black: Pa7,b7,c6,e6,f7,g7,h6 Na6,b6 (!) Bc5 Ra8,f8 Qe7 Kg8 >>> >>>White was pleased to find the continuation: >>> >>>18 f5! hxg5 >>>19 f6 gxf6 >>>20 Ne4 Rfd8 >>>21 Nxf6+ Qxf6 >>>22 Rxf6 Nd7 >>>23 Bxg5 Nxf6 >>>24 Bxf6 Kf8 >>>25 Qh5 Ke8 >>>26 Rf1 resigns >>> >>>It would seem obvious that White is winning within a 9-ply search, >>>so why doesn't any program find it? (You're of course welcome to >>>see for yourself of it's 'legal', but I can't find any get-out for Black.) >> >>The reason no program can find it is that 9 plies is insufficient to see the >>threat. >> >>My program favored Nfe4 through ply 11. I made the moves 18. f5 hxg6 and let my >>program think. After 12 plies and an eternity I got tired of waiting and stopped >>it; the pv was 19. h4 g4 20. Qxg4 , with a score of 0.00. >> >>James > >After 9 plies (played) Genius4 reckons itself at worse than -5.0 (Black's >already had >to give up his queen, rather more than a "threat"). How come a 9+ ply search >doesn't find this? > >Francis Yes, but if black *doesn't* give up his queen, it takes white several more plies to mate. To play 18. f5, White must see that he can mate (in more than 9 plies) if Black doesn't give up the queen. James
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.