Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Why do engines blunder here, playing 37.h5??

Author: Martin Giepmans

Date: 22:57:43 11/22/02

Go up one level in this thread


On November 23, 2002 at 01:01:32, Mike S. wrote:

>From a game Shredder 5.32 - Ruffian 1.0.1, 40/1h P3/700:
>
>[D]5k2/pp2r2p/2p2qp1/2Pn1p2/PPN1p2P/3rP3/5PP1/1QR1R1K1 w - - 0 1
>
>Analysis by Shredder 5.32:
>
>37.Dxd3 exd3
>  -+  (-4.29)   Tiefe: 1/2   00:00:00
>37.h5 gxh5 38.b5 h4 39.bxc6 Dxc6 40.Db2 Df6 41.Dxf6+ Sxf6 42.Sd6 Td5
>  =  (0.11)   Tiefe: 1/2   00:00:00
>  =  (-0.12)  Tiefe: 15/30   00:31:54  508626kN
>
>Analysis by Chess Tiger 14.0:
>
>37.h5 Tc7
>  ³  (-0.64)   Tiefe: 8   00:00:00  138kN
>  ³  (-0.45)   Tiefe: 15  00:10:29  212301kN
>
>Analysis by Fritz 7:
>
>37.h5 gxh5 38.b5 cxb5 39.axb5 Tg7 40.b6 axb6 41.Sxb6 Dg6 42.g3 h4 43.Sxd5
>  ³  (-0.56)   Tiefe: 7/23    00:00:00  160kN
>  µ  (-0.78)   Tiefe: 10/31   00:00:06  3467kN
>37.Ted1 Txd1+ 38.Txd1 Dxh4 39.b5 f4 40.exf4 Dxf4 41.Db2 Kg8 42.Sd6 Dg4 43.Dd4
>Dh4
>  µ  (-0.75)   Tiefe: 10/31   00:00:06  3789kN
>  ³  (-0.53)   Tiefe: 12/33   00:00:16  10341kN
>37.h5 gxh5 38.b5 Tg7 39.Db2 Dg5 40.g3 h4 41.De5 Sc7 42.b6 axb6 43.cxb6
>  ³  (-0.50)   Tiefe: 12/34   00:00:32  20236kN
>  µ  (-1.09)   Tiefe: 15/43   00:16:11  611858kN
>
>Why do many engines waste the h-pawn here, with no obvious reason? Weakening f5
>has no perspective; as it turns out Nc4-d6 can't harm Black. The open g-file is
>much to Black's favour.
>
>Usually I don't see this type of blunder from top engines. Is there something
>"computer-specific" which causes 37.h5?? here? The normal move is 37.g3, which
>certainly isn't worse than opening an attack file against the own king (?!).
>
>Thanks,
>M.Scheidl
>
>
>P.S. the game:
>
>[Site "40/1h P3/700"]
>[Date "2002.11.14"]
>[Round "1.5"]
>[White "Shredder 5.32"]
>[Black "Ruffian 1.0.1"]
>[Result "0-1"]
>[ECO "A21"]
>[PlyCount "84"]
>
>1. c4 e5 2. Nc3 c6 3. Nf3 Qc7 4. d4 d6 5. e3 Be7 6. Bd3 Nf6 7. O-O O-O 8. h3
>Na6 9. a3 Bd7 10. Qc2 Rfe8 11. b4 Rad8 12. dxe5 dxe5 13. Ne4 g6 14. Bb2 Nxe4
>15. Bxe4 Bf8 16. Rad1 Bg7 17. Qb3 Be6 18. Ng5 Bc8 19. Bb1 Qe7 20. Nf3 e4 21.
>Nd2 f5 22. Bxg7 Qxg7 23. c5+ Be6 24. Nc4 Nc7 25. Qc2 Nb5 26. Ba2 Qc3 27. Qe2
>Rd3 28. Rc1 Qg7 29. Na5 Bxa2 30. Qxa2+ Kf8 31. Nc4 Re7 32. h4 Red7 33. a4 Nc3
>34. Qc2 Nd5 35. Qb1 Qf6 36. Rfe1 Re7 37. h5 gxh5 38. Nd6 Rg7 39. Nc4 f4 40. Qb2
>Qg5 41. f3 fxe3 42. f4 Nxf4 0-1


The problem is maybe that the engines "fear" that black will answer g3 (the move
that you propose) with the thematic f4.
When I feed my engine the position after g3 f4 it doesn't see a clear win for
black (at least not within a few minutes) but it isn't happy with the situation.

Martin



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.