Author: Günther Simon
Date: 10:21:29 10/13/05
Go up one level in this thread
On October 13, 2005 at 12:58:43, Dagh Nielsen wrote:
>I have used Fruit 2.1 and Fruit 2.2 for many games on playchess now. And I have
>seen many times when Fruit will happily enter an endgame an exchange down but a
>pawn up + maybe some activity. It generally evaluates these positions as equal
>before entering them, but then, of course, 5-10 moves later the rook shows its
>superiority.
>
>Here's an example
>
>[Event "Rated game, 3m + 0s"]
>[Site "Engine Room"]
>[Date "2005.10.13"]
>[Round "?"]
>[White "Cerberus1979, Fritz 9"]
>[Black "Flying Saucers, Fruit 2.2"]
>[Result "1-0"]
>[ECO "C78"]
>[WhiteElo "2554"]
>[BlackElo "2617"]
>[PlyCount "103"]
>[EventDate "2005.03.23"]
>[TimeControl "180"]
>
>1. e4 {0} e5 {0} 2. Nf3 {0} Nc6 {0} 3. Bb5 {0} a6 {0} 4. Ba4 {0} Nf6 {0} 5. O-O
>{0} b5 {(Le7) 0} 6. Bb3 {0} Bb7 {0} 7. Re1 {0} Bc5 {0} 8. c3 {0} d6 {0} 9. d4 {
>0} Bb6 {0} 10. Bg5 {0} h6 {0} 11. Bh4 {0} g5 {0} 12. Bg3 {0} O-O {0} 13. dxe5 {
>0} Nxe5 {0} 14. Nxe5 {0} dxe5 {0} 15. Qf3 {0} Qe7 {0} 16. Qf5 {0} Kg7 {0} 17.
>Bxe5 {0} Rad8 {0} (17... Rfe8 18. Bxf6+ Qxf6 19. Qxf6+ Kxf6 20. Nd2) 18. Bxf6+
>{0} Qxf6 {0} 19. Qxf6+ {0} Kxf6 {0} 20. e5+ {0.11/10 4} Kg7 {(Ke7) 0.13/13 1}
>21. Re2 {0.42/10 4} Rfe8 {0.03/13 3} 22. a4 {0.42/11 3} Bd5 {(c5) 0.08/13 6} (
>22... c5 23. Na3 c4 24. Bc2 Bc8 25. axb5 axb5 26. Nxb5 Bg4 27. Ree1 Rd2 28. Nd6
>Bxf2+ 29. Kf1 Rb8 30. Bd1 Bxe1 31. Bxg4 Rf2+ 32. Kxe1 Rxg2) 23. Nd2 {0.17/11 5}
>Bxb3 {0.03/15 2} 24. Nxb3 {0.20/12 2} bxa4 {0.07/16 7} 25. Nd2 {0.19/12 0} Rxe5
>{(Td5) 0.07/15 5} 26. Rxe5 {0.51/12 4} Rxd2 {0.00/15 1} 27. Rf5 {0.43/12 3}
>Rxb2 {-0.12/15 3} 28. g4 {0.26/13 5} Rb5 {-0.09/16 6} 29. Rxa4 {0.26/13 0} Rxf5
>{-0.03/18 6} 30. gxf5 {0.45/15 0} a5 {0.00/19 4} 31. c4 {0.26/14 0} Kf6 {
>0.00/19 6} 32. Kg2 {0.36/14 0} c6 {0.00/19 15} 33. Ra3 {0.39/15 0} Kxf5 {
>-0.02/17 5} 34. Rh3 {0.46/13 0} Kg6 {-0.16/17 4} 35. Rb3 {0.47/13 0} Bc5 {
>(Ld4) 0.00/17 3} 36. Rb8 {0.56/14 5} a4 {0.27/17 0} 37. Rc8 {0.64/13 4} f6 {
>(Kg7) 0.35/18 0} 38. Rxc6 {0.81/13 4} Be7 {(Lb4) 0.62/17 1} 39. Kf3 {1.21/13 4}
>Kf5 {0.89/17 0} 40. Ra6 {1.13/13 2} a3 {1.04/17 3} 41. Ra5+ {1.41/13 2} Ke6 {
>1.13/18 2} 42. Ke4 {1.43/12 0} f5+ {(Ld6) 1.12/17 3} 43. Rxf5 {1.52/13 5} Kd7 {
>1.09/19 0} 44. Kd3 {1.84/11 3} Bd6 {1.21/15 4} 45. h3 {2.09/13 1} Be7 {
>(Lb4) 1.39/15 4} 46. Kc3 {2.26/13 4} Bd6 {(h5) 1.75/16 5} 47. Kb3 {2.71/13 4}
>Ke6 {(Le7) 2.11/16 2} 48. Ra5 {2.85/14 3} Bf8 {2.30/17 3} 49. f3 {2.92/13 1}
>Be7 {2.34/16 2} 50. Ra6+ {3.01/13 0} Ke5 {3.00/17 3} 51. Rxh6 {3.13/13 0} Kf4 {
>3.14/17 4} 52. Rh7 {Flying Saucers,Fruit 2.2 resigns 3.30/13 0} 1-0
>
>After 22. a4 Fruit chooses to enter an endgame an exhange down, resulting in
>this position at move 30:
>
>[D] 8/2p2pk1/pb5p/5Pp1/R7/2P5/5P1P/6K1 b - - 0 30
>
>It evaluates this position as dead equal, which may be right, but already at
>move 32 it makes a positional blunder with c7-c6 and loses rather quickly.
>
>This is just one example, I have seen many others. I would like to address it,
>maybe one could make a rook worth more in the settings, but that would disrupt
>the evaluation of rook + pawn vs. two minor pieces.
>
>In general, I must say I am very happy with Fruit as engine, and also its
>endgame play is ok, superior to Fritz 8 at least. But I think there is room for
>quite a lot of improvement in Fruit's endgame evaluations, making it an even
>stronger engine and ready to challenge Schredder 10 once that time comes.
Well S9.1 makes the same 'blunder' at least in a fast game...
What speed has your machine? I did this analysis on my slow machine P3 650,
but as it still holds c6/depth 21 as best after 47 sec, it will hold it for at
least 14 sec on a P4 2.67 (ratio 3.3:1 measured here, P3s are relatively
faster...)
I guess people should stop concluding too much from bullet/fast blitz sudden
death tcs, this just leads often to wrong conclusions like here.
(the result was just a random result due to thinking times of 0-3 sec from
an equal game at around move 30.)
Guenther
[D]8/2p2p2/1b3k1p/p4Pp1/R1P5/8/5PKP/8 b - -
Engine: Shredder 9 UCI (128 MB)
von Stefan Meyer-Kahlen
13/19 0:01 0.00 32...c6 33.c5 Ld8 34.Td4 Le7 35.Ta4 Ld8
36.Td4 Le7 37.Ta4 Ld8 (242.937) 160
14/18 0:02 0.00 32...c6 33.c5 Ld8 34.Td4 Le7 35.Ta4 Ld8
36.Td4 Le7 37.Ta4 Ld8 (348.159) 165
15/23 0:03 +0.25++ 32...c6 33.c5 (599.154) 176
15/23 0:03 +0.25 32...c6 33.c5 Lc7 34.Td4 Kxf5 35.Td7 (654.667) 177
16/23 0:04 +0.22 32...c6 33.c5 Lc7 34.Td4 Kxf5 35.Td7 Le5
36.Txf7+ Ke6 37.Ta7 Lc3 38.Kf3 Kd5
39.Ke2 Kxc5 40.Kd3 Kb4 (883.592) 182
17/23 0:06 +0.27 32...c6 33.c5 Lc7 34.Td4 Kxf5 35.Td7 Le5
36.Txf7+ Ke6 37.Ta7 Lc3 38.Kf3 Kd5
39.Ke2 Kxc5 40.Kd3 Lb4 41.Ta6 (1.224.539) 186
18/25 0:08 +0.02-- 32...c6 33.c5 (1.512.433) 188
18/27 0:11 0.00 32...c6 33.c5 Ld8 34.Td4 Le7 35.Ta4 Ld8
36.Td4 Le7 37.Ta4 Ld8 (2.212.314) 192
19/25 0:16 0.00 32...c6 33.c5 Ld8 34.Td4 Le7 35.Ta4 Ld8
36.Td4 Le7 37.Ta4 Ld8 (3.254.457) 195 TB:8
20/31 0:23 0.00 32...c6 33.c5 Ld8 34.Td4 Le7 35.Ta4 Ld8
36.Td4 Le7 37.Ta4 Ld8 (4.528.860) 196 TB:14
21/33 0:39 -0.15 32...c6 33.c5 Ld8 34.Ta1 Kxf5 35.Te1 Lf6
36.Te8 Lc3 37.Tc8 Ke5 38.Txc6 Kd4
39.Txh6 Kxc5 40.Ta6 (7.823.098) 199 TB:36
Bester Zug: c7-c6 Zeit: 0:46.497 min K/s: 199.297 CPU 97.4% K/s(1CPU):
204.617 Knoten: 9.262.763 TB: 48
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.