Author: Laurence Chen
Date: 16:14:13 03/31/99
Go up one level in this thread
On March 31, 1999 at 14:02:35, Jon Dart wrote: >Here is an interesting position from a game my program >played on FICS recently: > >r1r2k2/1p4pQ/1n1bp2p/p2n1p2/P2P2Nq/1B5P/1P3PP1/R1BR2K1 w - f6 > >White (the human player) played Re1, leaving the Knight hanging, >and won, after fxg4 Rxe6 Rxc1+ Rxc1 Qg5 Rce1 gxh3 Re8+ Kf7. >Very impressive, especially in a blitz game! However, my program >did not find the best defense : fxg4 Rxe6 gxh3, etc. > >Crafty 16.5 likes Re1, but it takes a long time to see it, and >even then, with the amount of time I gave it, assigned it only >a small plus score. I'm not sure it's enough better than the >alternatives to be a "best move". Here is the output: > > 11 2:49 -0.37 1. Ne3 Qh5 2. Nxd5 Nxd5 3. Bd2 Kf7 > 4. Bxd5 exd5 5. Bxh6 Bf8 6. b4 Qxh6 > 7. Qxf5+ Qf6 8. Qxd5+ > 11 5:37 ++ 1. Re1!! > 11 6:20 0.10 1. Re1 fxg4 2. Rxe6 gxh3 3. Qf5+ Nf6 > 4. Rxd6 h2+ 5. Kh1 Nc4 6. Rd7 b6 7. > d5 > 11-> 7:39 0.10 1. Re1 fxg4 2. Rxe6 gxh3 3. Qf5+ Nf6 > 4. Rxd6 h2+ 5. Kh1 Nc4 6. Rd7 b6 7. > d5 > 12 11:31 0.08 1. Re1 fxg4 2. Rxe6 gxh3 3. Qf5+ Nf6 > 4. Rxd6 h2+ 5. Kh1 Nc4 6. Rd7 b6 7. > d5 Qg4 > 12-> 15:35 0.08 1. Re1 fxg4 2. Rxe6 gxh3 3. Qf5+ Nf6 > 4. Rxd6 h2+ 5. Kh1 Nc4 6. Rd7 b6 7. > d5 Qg4 > 13 26:30 0.06 1. Re1 fxg4 2. Rxe6 gxh3 3. Qf5+ Nf6 > 4. Rxd6 Nc4 5. Rd7 h2+ 6. Kh1 <HT> > time=33:20 cpu=99% mat=0 n=344343893 fh=94% nps=172105 > ext-> checks=20911812 recaps=1000315 pawns=174053 1rep=1446615 > predicted=0 nodes=344343893 evals=34920023 > endgame tablebase-> probes done=0 successful=0 plimit=17 > hashing-> trans/ref=19% pawn=4% used=99% > >Here is the whole game: > >[Event "?"] >[Site "FICS"] >[Date "1999.03.30"] >[Round "?"] >[White "bielg"] >[Black "Arasan 5.1"] >[Result "1-0"] >[ECO "D18"] >[WhiteELO "2227"] >[BlackELO "2270"] >[TimeControl "120+12"] > >1. Nf3 d5 2. d4 c6 3. c4 Nf6 4. Nc3 dxc4 5. a4 Bf5 6. e3 e6 7. Bxc4 >Bb4 8. O-O Nbd7 9. h3 O-O 10. Qb3 c5 11. Nh4 Bxc3 12. Nxf5 Bb4 >13. Ng3 Nb6 14. Bd3 a5 15. Rd1 cxd4 16. exd4 Nfd5 17. Ne4 Qh4 18. Ng5 >h6 19. Bh7+ Kh8 20. Nf3 Qf6 21. Bc2 Kg8 22. Qd3 Rfc8 23. Qh7+ Kf8 >24. Ne5 Bd6 25. Ng4 Qh4 26. Bb3 f5 27. Re1 fxg4 28. Rxe6 Rxc1+ >29. Rxc1 Qg5 30. Rce1 gxh3 31. Re8+ Kf7 1-0 Is this a trick question? What's the best move? The way I see this position is White has a positional won game on move 26, and Arasan completely mis-evaluated the position on move 21. Bc2 and played the ugly bad move Kg8?? Fritz 5.32 prefers 21. ... Qe7 which leads to an equal chances game although I prefer Black's position after 21. ... Qe7. The human master played a very strong move 27. Re1! This move has a lot of sting and it was played on positional grounds. Although IMO at this point the game has already been decided and White has a positional won game, and there's no defense for Black. There are many other alternatives moves which will lead to a won game for White. Whether these alternatives are better than the actual move played remains to be open for discussion. Note the position before white's move 27. Re1, Black has 5 pieces sitting in the Queenside, and the only piece left to defend the kingside is a lonely Queen. White has two Bishops pointing their gun to the kingside, and a knight which was sacrificed in order to open lines, the open e-file for the rooks, and diagonals for the bishops. So in another words White has all his pieces attacking the kingside, and Black's king is not able to run to the Queenside for shelter where all his army lies. Black is outnumbered in the kingside and therefore the sad defeat because he is unable to bring his army sitting in the Queenside to help on the defense of his poor king. :) Laurence
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.