Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: pawn endgame position

Author: Robert Hyatt

Date: 12:59:50 01/22/01

Go up one level in this thread


On January 22, 2001 at 09:36:53, Steffen Jakob wrote:

>Hossa played a blitz game with Amateur some minutes ago where he got a won
>endgame position. Both sides had a queen and pawns. Hossa was 3 pawns up. Then
>there was the opportunity to exchange the queens resulting in a pawn endgame
>position where Hossa had 3 connected advanced passed pawns and Amateur had no
>passer at all. You can imagine that Hossa was very happy to exchange in this
>line. Unfortunately the 3 passers werent able to force a promotion without the
>help of their king who was far away. On the other hand black could create a
>passer which could not be stopped by white. So in the end Hossa lost this
>"deadly" won position :-)
>
>I am interested if "your" engine could resist to play 46.Qe6+ in a blitz game.
>
>[D]6k1/2pq4/5Q2/p2PPP2/1p1P4/8/P6P/7K w
>
>Best wishes,
>Steffen.
>
>[Event "ICS rated blitz match"]
>[Site "204.178.125.65"]
>[Date "2001.01.22"]
>[Round "-"]
>[White "Hossa"]
>[Black "Amateur"]
>[Result "0-1"]
>[WhiteElo "2508"]
>[BlackElo "2548"]
>[TimeControl "300+2"]
>
>1. d4 d5 2. c4 e6 3. Nc3 Nf6 4. cxd5 exd5 5. Bg5 Be7 6. Nf3 h6 7. Bh4 O-O
>8. e3 Nc6 9. Be2 Be6 10. O-O Ne4 11. Bxe7 Qxe7 12. Qb3 Nxc3 13. bxc3 Rab8
>14. Bd3 Rfd8 15. Qc2 Bg4 16. Nd2 Qg5 17. f4 Qf6 18. Nf3 Ne7 19. Ne5 Bh5 20.
>Bh7+ Kh8 21. g4 Bg6 22. Bxg6 Nxg6 23. c4 Nxe5 24. fxe5 Qe6 25. cxd5 Qxg4+
>26. Kh1 Rf8 27. Rf4 Qc8 28. Rg1 Qd7 29. e4 Qe7 30. Qc1 b5 31. Rf3 b4 32.
>Rfg3 Rg8 33. R1g2 a5 34. Qf4 Rbd8 35. Rh3 f5 36. Rxh6+ gxh6 37. Qxh6+ Qh7
>38. Qf6+ Rg7 39. Qxd8+ Qg8 40. Qf6 Qf7 41. Rxg7 Qxg7 42. Qh4+ Kg8 43. exf5
>Qd7 44. Qg5+ Kf8 45. Qf6+ Kg8 46. Qe6+ Qxe6 47. fxe6 Kf8 48. d6 cxd6 49.
>exd6 a4 50. d7 Ke7 51. Kg2 b3 52. axb3 axb3 53. d5 b2 54. Kf3 b1=Q 55. Kf4
>Qd3 56. Ke5 Qe3+ 57. Kf5 Qd4
>{Hossa resigns} 0-1


Here is what I get:

         (2)    8->   0.54   5.15   1. Qe6+ Qf7 2. Qg6+ Kf8 3. Qh6+ Kg8
                                    4. Qg5+ Qg7 5. Qxg7+ Kxg7 6. h4 a4
                                    7. f6+ Kf7 8. h5
                9     0.66   5.49   1. Qe6+ Qf7 2. Qg6+ Kf8 3. Qh6+ Kg8
                                    4. Qg5+ Qg7 5. Qxg7+ Kxg7 6. h4 Kf7
                                    7. h5 Ke7 8. h6
         (2)    9->   0.94   5.49   1. Qe6+ Qf7 2. Qg6+ Kf8 3. Qh6+ Kg8
                                    4. Qg5+ Qg7 5. Qxg7+ Kxg7 6. h4 Kf7
                                    7. h5 Ke7 8. h6
               10     1.25     ++   1. Qe6+!!
               10    11.21   6.34   1. Qe6+ Qxe6 2. fxe6 Kf8 3. d6 cxd6
                                    4. exd6 Ke8 5. d5 Kd8 6. h4 Ke8 7.
                                    h5
               10->  11.91   6.34   1. Qe6+ Qxe6 2. fxe6 Kf8 3. d6 cxd6
                                    4. exd6 Ke8 5. d5 Kd8 6. h4 Ke8 7.
                                    h5
               11    12.17     --   1. Qe6+
               11    12.65   3.57   1. Qe6+ Qxe6 2. fxe6 Kf8 3. d6 cxd6
                                    4. exd6 Ke8 5. d7+ Kd8 6. Kg2 a4 7.
                                    d5 b3 8. e7+ Kxe7 <HT>
               11    17.67     ++   1. e6!!
               11    38.49   4.49   1. e6 Qxd5+ 2. Kg1 Qe4 3. Qd8+ Kg7
                                    4. Qxc7+ Kh6 5. Qe5 Qb1+ 6. Kg2 Qxa2+
                                    7. Kg3 Qd2 8. e7 Qg5+ 9. Kf2 Qh4+ 10.
                                    Ke2 Qg4+ 11. Kd3
               11->  43.08   4.49   1. e6 Qxd5+ 2. Kg1 Qe4 3. Qd8+ Kg7
                                    4. Qxc7+ Kh6 5. Qe5 Qb1+ 6. Kg2 Qxa2+
                                    7. Kg3 Qd2 8. e7 Qg5+ 9. Kf2 Qh4+ 10.
                                    Ke2 Qg4+ 11. Kd3


It is very likely it would see that in a blitz game.  Pondering effects
would probably reduce the total time by a factor of 2 or so,
and 12 seconds is easily doable in blitz...

This is a case where I don't try to statically evaluate too much.  The black
candidate at b-whatever gets a good bonus, but not as much as the passers do
for white, making this look good for white.  My connected passer code knows
that in K+P endings, if the connected passers are blockaded by the king they
are weak, but not weak enough.  I also do not do 'pawn race' tests for
candidates as that is complicated.  It would certainly fix this case as the b
pawn is a problem.  As is white's horrible king position.



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.