Author: Ricardo Gibert
Date: 20:26:19 03/07/04
Go up one level in this thread
On March 07, 2004 at 21:37:26, Mike Byrne wrote:
>I have often heard that chess programmers like to steer their programs away from
>the King's Indian -- any thoughts on that why it may be true because it seems to
>be very solid for Black with most of best players playing it.
I remember once watching Crafty play a King's Indian on ICC against another
chess program. It was the classical variation.
After leaving book, Crafty decided it needed to develop its White squared B on
c8, so it played b6 and Bb7. Against a strong human, this is enough for a
strategically lost position in the classical variation. IIRC, it later even
exchanged off this B, which does not help things though technically it made no
difference at that point.
What neither program understood was that Black's white squared B is a critical
piece for Black in the classical vaiation. It usually just stays on c8, since it
is already on its best square there. This piece is destined to sac' itself on
h6. Without this possibility, Black's attack on the K-side slows down too much
or is doomed to fail entirely.
The other chess program did not know how to attack on the Q-side and shifted
about, but nevertheless managed a little progress, but eventually in it's
meanderings, it managed to box it's Q in corner on a3, which put it out of play.
Crafty managed to take advantage of this by sac'ing to open up the K-side to
manage some interesting counterplay and went on to win.
The game was quite comical to watch. Both programs were clueless and it was
evident that the King's Indian is an opening that programs should avoid playing
against humans. At least the more closed variations like the classical variation
should be avoided.
>
>Coincidently, I recently picked up Gligoric's Kings Indian Defence: Mar Del
>Plata Variation - Svetozar is an excellent chesswriter and I enjoyed his writing
>style so much I also picked up his "I Play Against Pieces". His style does not
>overwhelm you with variations but actually uses the words to express the ideas
>behind his thinking. He is clearly one of the best chess writers.
>
>
>Below is game against Fischer that he won. This yet another game that I had to
>correct from Chesslib database. The game actually ended at move 44. - the
>database shows Fischer getting mated which simply did not happen. I sprinked
>juts a couple the annotations from Gligoric's book to show you his style. The
>game in the book is much more heavily annotated.
>
>
>[Event "Warna"]
>[Site "Warna"]
>[Date "1962.??.??"]
>[Round "?"]
>[White "Fischer, Robert J"]
>[Black "Gligoric, Svetozar"]
>[Result "0-1"]
>[ECO "B80"]
>[PlyCount "88"]
>[EventDate "1962.??.??"]
>[Source "www.ChessliB.no"]
>[SourceDate "2003.02.24"]
>
>1. e4 c5 2. Nf3 d6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 a6 6. g3 e6 7. Bg2 Be7 8. O-O
>O-O 9. f4 Qc7 10. g4 Nc6 11. Nxc6 bxc6 12. g5 Nd7 13. f5 Re8 {Trying to orgnaize
>some kind of defense against the threat of f5-f6 (Gligoric)}14. Kh1 Bf8 15.
>Bf4 Ne5 16. f6 g6 17. h4 a5 18. h5 Ba6 19. Re1 Qb6 20. hxg6 fxg6 21. Bxe5
>{Removing the knight that might come to f7 (Gligoric)} dxe5
>22. Qf3 Ra7 23. Bf1 Rf7 24. Bxa6 Qxa6 25. Qg3 Qb6 26. Qxe5 Qxb2 27. Rad1 h6 28.
>Re3 Bb4 29. gxh6 Qxc2 30. Rg1 Kh7 31. Qg3 Rg8 32. e5 Bxc3 33. Rxc3 Qe4+ 34. Rg2
>Rd8 35. Re3 Qb1+ 36. Kh2 Rd1 37. Qg4 Rh1+ 38. Kg3 Qc1 {
>
>[d]8/5r1k/2p1pPpP/p3P3/6Q1/4R1K1/P5R1/2q4r w - - 0 39
>Although Ree2 might have offered tougher resistance. (Byrne)
>
>39. Re4? (39. Ree2 Rd7 40.
>Qe4 Rh5 41. Rc2 Qa3+) 39... Rd7 40. Qe2 Qg5+ 41. Qg4 Rd3+ 42. Kf2 Rd2+ 43. Kg3
>Rxg2+ 44. Kxg2 Qc1 0-1
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.