Not logged inRybka Chess Community Forum
Up Topic Rybka Support & Discussion / Rybka Discussion / Rybka 3 cannot find an obvious draw
- - By domelius [de] Date 2010-02-16 02:54 Edited 2010-02-16 03:51
Hi people,

a few words about me first: I am a weak player from Bulgaria, my current national rating is 2120, I dont have international Elo (must be similar if I had one).
I played a Blitz game against Rybka 3 at full strenght, without using any help and without taking back any moves. Rybka had 1 min for the whole game, I had a human Bonus of 1 hour. I was very happy to achieve a draw thanks to a weakness into the position valuation of the engine. Here I paste the game moves:

[Event "Blitz:1'"]
[Site "?"]
[Date "2010.02.16"]
[Round "?"]
[White "Rybka 3 1-cpu 32-bit"]
[Black "Dimov, Dimcho"]
[Result "1/2-1/2"]
[ECO "B22"]
[Annotator ",Dimcho"]
[PlyCount "100"]
[EventDate "2010.??.??"]
[TimeControl "60"]

{24MB, Fritz11.ctg, DIMOV} 1. e4  c5  2. c3  d5  3. exd5  Qxd5
4. d4  Nf6  5. Nf3  g6  6. Qb3  Qxb3  7. axb3  cxd4  8. Nxd4  Bg7  9. Nb5  Na6  10. Rxa6  bxa6  11. Nc7+
Kd7  12. Nxa8  Bb7  13. f3  Rxa8  14. Na3  e6  15. Nc4  Nd5  16. Na5  Bc8  17. b4  Rb8  18. Bd3  Nc7  19.
Be3  Nb5  20. Kf2  Bb7  21. Nxb7  Rxb7  22. Ra1  Nxc3  23. bxc3  Bxc3  24. Rxa6  Bxb4  25. g3  f6  26. Bc4
Bd6  27. Rxa7  Rxa7  28. Bxa7  e5  29. Bd5  f5  30. Be3  f4  31. gxf4  exf4  32. Bd4  g5  33. Ke2  h5  34. Bf7
g4  35. Bxh5  gxf3+  36. Kxf3  Ke7  37. h4  Ke6  38.  Ke4  Be7  39. Bg4+  Kf7  40. h5  Bf6  41. Bxf6  Kxf6 
42. Kxf4  Kg7  43. Kg5  Kh8  44. Kg6  Kg8  45. h6  Kh8  46. Bf5  Kg8  47. Bb1  Kh8  48. Bd3  Kg8  49. Be4  Kh8
50. Bb1  Kg8  1/2-1/2

This position is a dead draw, and the program still gives a +6.00 for white. The draw is achieved after 41. Bxf6, which is a blunder. The program should also avoid to exchange the f and g pawns.  Fritz 11 on the other hand refuses correctly to exchange the bishops in this position and after 41. Bxf6 sees the draw immediately.

And here a similar position:

Black to move and draw:


Rybka cannot find the line 1...Ra1+ 2. Rf1 Rxf1+ 3. Kxf1 Bh3!!. The move 3... Bh3!! is the only move which saves the game for back, and is easily found by Fritz 11 and Shredder 12, but not by Rybka.

I am writing this statement not to boast about my draw with Rybka, I just want to give you a tip for improvement for the next versions, if you havent already noticed this weakness.

Best regards,

D.Dimov
Parent - By vdragan (***) [us] Date 2010-02-16 03:50
noticed loooong time ago :)
Parent - By Gaмßito (****) [cr] Date 2010-02-16 03:52 Edited 2010-02-16 04:04
Yes, this is already a very known topic.

The wrong bishop knowledge was removed after Rybka 2.1. Then, Vas promised to return this knowledge again for Rybka 3 but he forgot it. And many months ago, he told us that this problem was already fixed, so there will be no problems with Rybka 4 for these positions. (If Rybka 4 is released. But I doubt it, ever more and more. :-)).

r7/8/8/5bk1/8/5B2/5RPP/6K1 b - - 0 1


Analysis by Rybka 1.2n 32-bit:

1...Ra1+ 2.Rf1 Ra3 3.g3 Bh3
  +/-  (1.05)   Depth: 6   00:00:00  15kN
1...Ra1+ 2.Rf1 Ra4 3.g3 Bh3 4.Re1
  +/-  (1.05)   Depth: 7   00:00:00  34kN
1...Ra1+ 2.Rf1 Ra4 3.g3 Bh3 4.Re1 Kf6
  +/-  (1.10)   Depth: 8   00:00:00  54kN
1...Ra1+ 2.Rf1 Ra3 3.Kf2 Rc3 4.Re1 Rc2+ 5.Re2 Rc4
  +/-  (1.15)   Depth: 9   00:00:01  104kN
1...Ra1+ 2.Rf1 Rxf1+ 3.Kxf1 Bh3 4.Kf2 Bxg2 5.Bxg2 Kf6 6.h4 Kg7 7.h5
  =  (0.00)   Depth: 10   00:00:02  244kN
1...Ra1+ 2.Rf1 Rxf1+ 3.Kxf1 Bh3 4.Kf2 Bxg2 5.Bxg2 Kf6 6.h4 Kg7 7.h5 Kh7
  =  (0.00)   Depth: 11   00:00:02  303kN
1...Ra1+ 2.Rf1 Rxf1+ 3.Kxf1 Bh3 4.Kf2 Bxg2 5.Bxg2 Kf6 6.h4 Kg7 7.h5 Kh7 8.Ke3
  =  (0.00)   Depth: 12   00:00:03  414kN
1...Ra1+ 2.Rf1 Rxf1+ 3.Kxf1 Bh3 4.Kf2 Bxg2 5.Bxg2 Kf6 6.h4 Kg7 7.h5 Kh7 8.Be4+
  =  (0.09)   Depth: 13   00:00:04  663kN
1...Ra1+ 2.Rf1 Rxf1+ 3.Kxf1 Bh3 4.Kf2 Bxg2 5.Bxg2 Kf6 6.h4 Kg7 7.h5 Kh7 8.Be4+
  =  (0.09)   Depth: 14   00:00:05  1033kN
1...Ra1+ 2.Rf1 Rxf1+ 3.Kxf1 Bh3 4.Kf2 Bxg2 5.Bxg2 Kf6 6.h4 Kg7 7.h5 Kh7 8.Be4+
  =  (0.09)   Depth: 15   00:00:06  1452kN
1...Ra1+ 2.Rf1 Rxf1+ 3.Kxf1 Bh3 4.Kf2 Bxg2 5.Bxg2 Kf6 6.h4 Kg7 7.h5 Kh7 8.Be4+
  =  (0.09)   Depth: 16   00:00:08  2041kN
1...Ra1+ 2.Rf1 Rxf1+ 3.Kxf1 Bh3 4.Kf2 Bxg2 5.Bxg2 Kf6 6.h4 Kg7 7.h5 Kh7 8.Be4+
  =  (0.09)   Depth: 17   00:00:10  2872kN
1...Ra1+ 2.Rf1 Rxf1+ 3.Kxf1 Bh3 4.Kf2 Bxg2 5.Bxg2 Kf6 6.h4 Kg7 7.h5 Kh7 8.Be4+
  =  (0.09)   Depth: 18   00:00:13  3978kN
1...Ra1+ 2.Rf1 Rxf1+ 3.Kxf1 Bh3 4.Kf2 Bxg2 5.Bxg2 Kf6 6.h4 Kg7 7.h5 Kh7 8.Be4+
  =  (0.09)   Depth: 19   00:00:17  5374kN
1...Ra1+ 2.Rf1 Rxf1+ 3.Kxf1 Bh3 4.Kf2 Bxg2 5.Bxg2 Kf6 6.h4 Kg7 7.h5 Kh7 8.Be4+
  =  (0.09)   Depth: 20   00:00:23  7182kN
1...Ra1+ 2.Rf1 Rxf1+ 3.Kxf1 Bh3 4.Kf2 Bxg2 5.Bxg2 Kf6 6.h4 Kg7 7.h5 Kh7 8.Be4+
  =  (0.09)   Depth: 21   00:00:32  9649kN
1...Ra1+ 2.Rf1 Rxf1+ 3.Kxf1 Bh3 4.Kf2 Bxg2 5.Bxg2 Kf6 6.h4 Kg7 7.h5 Kh7 8.Be4+
  =  (0.09)   Depth: 22   00:00:42  12839kN

(AMD, 15.02.2010)

6k1/8/6KP/8/8/8/8/1B6 w - - 0 1


Analysis by Rybka 1.2n 32-bit:

1.h7+ Kh8 2.Kg5 Kg7 3.h8Q+ Kxh8 4.Kf6
  =  (0.09)   Depth: 6   00:00:00  11kN
1.h7+ Kh8 2.Kg5 Kg7 3.h8Q+ Kxh8 4.Kf6 Kg8
  =  (0.09)   Depth: 7   00:00:00  21kN
1.h7+ Kh8 2.Kg5 Kg7 3.h8Q+ Kxh8 4.Kf6 Kg8 5.Ba2+ Kh7
  =  (0.09)   Depth: 8   00:00:00  37kN
1.h7+ Kh8 2.Kg5 Kg7 3.h8Q+ Kxh8 4.Kf6 Kg8 5.Ba2+ Kh7 6.Bb3
  =  (0.09)   Depth: 9   00:00:00  61kN
1.h7+ Kh8 2.Kg5 Kg7 3.h8Q+ Kxh8 4.Kf6 Kg8 5.Ba2+ Kh7 6.Bb3 Kh8
  =  (0.09)   Depth: 10   00:00:00  96kN
1.h7+ Kh8 2.Kg5 Kg7 3.h8Q+ Kxh8 4.Kf6 Kg8 5.Ba2+ Kh7 6.Bb3 Kh8 7.Bc4 Kh7
  =  (0.09)   Depth: 11   00:00:00  143kN
1.h7+ Kh8 2.Kg5 Kg7 3.h8Q+ Kxh8 4.Kf6 Kg8 5.Ba2+ Kh7 6.Bb3 Kh8 7.Bc4 Kh7
  =  (0.09)   Depth: 12   00:00:00  207kN
1.h7+ Kh8 2.Kg5 Kg7 3.h8Q+ Kxh8 4.Kf6 Kg8 5.Ba2+ Kh7 6.Bb3 Kh8 7.Bc4 Kh7
  =  (0.09)   Depth: 13   00:00:01  286kN
1.h7+ Kh8 2.Kg5 Kg7 3.h8Q+ Kxh8 4.Kf6 Kg8 5.Ba2+ Kh7 6.Bb3 Kh8 7.Bc4 Kh7
  =  (0.09)   Depth: 14   00:00:01  396kN
1.h7+ Kh8 2.Kg5 Kg7 3.h8Q+ Kxh8 4.Kf6 Kg8 5.Ba2+ Kh7 6.Bb3 Kh8 7.Bc4 Kh7
  =  (0.09)   Depth: 15   00:00:01  538kN
1.h7+ Kh8 2.Kg5 Kg7 3.h8Q+ Kxh8 4.Kf6 Kg8 5.Ba2+ Kh7 6.Bb3 Kh8 7.Bc4 Kh7
  =  (0.09)   Depth: 16   00:00:02  711kN
1.h7+ Kh8 2.Kg5 Kg7 3.h8Q+ Kxh8 4.Kf6 Kg8 5.Ba2+ Kh7 6.Bb3 Kh8 7.Bc4 Kh7
  =  (0.09)   Depth: 17   00:00:02  903kN
1.h7+ Kh8 2.Kg5 Kg7 3.h8Q+ Kxh8 4.Kf6 Kg8 5.Ba2+ Kh7 6.Bb3 Kh8 7.Bc4 Kh7
  =  (0.09)   Depth: 18   00:00:03  1160kN
1.h7+ Kh8 2.Kg5 Kg7 3.h8Q+ Kxh8 4.Kf6 Kg8 5.Ba2+ Kh7 6.Bb3 Kh8 7.Bc4 Kh7
  =  (0.09)   Depth: 19   00:00:03  1458kN
1.h7+ Kh8 2.Kg5 Kg7 3.h8Q+ Kxh8 4.Kf6 Kg8 5.Ba2+ Kh7 6.Bb3 Kh8 7.Bc4 Kh7
  =  (0.09)   Depth: 20   00:00:04  1960kN
1.h7+ Kh8 2.Kg5 Kg7 3.h8Q+ Kxh8 4.Kf6 Kg8 5.Ba2+ Kh7 6.Bb3 Kh8 7.Bc4 Kh7
  =  (0.09)   Depth: 21   00:00:05  2251kN
1.h7+ Kh8 2.Kg5 Kg7 3.h8Q+ Kxh8 4.Kf6 Kg8 5.Ba2+ Kh7 6.Bb3 Kh8 7.Bc4 Kh7
  =  (0.09)   Depth: 22   00:00:05  2566kN
1.h7+ Kh8 2.Kg5 Kg7 3.h8Q+ Kxh8 4.Kf6 Kg8 5.Ba2+ Kh7 6.Bb3 Kh8 7.Bc4 Kh7
  =  (0.09)   Depth: 23   00:00:06  2905kN
1.h7+ Kh8 2.Kg5 Kg7 3.h8Q+ Kxh8 4.Kf6 Kg8 5.Ba2+ Kh7 6.Bb3 Kh8 7.Bc4 Kh7
  =  (0.09)   Depth: 24   00:00:06  3275kN
1.h7+ Kh8 2.Kg5 Kg7 3.h8Q+ Kxh8 4.Kf6 Kg8 5.Ba2+ Kh7 6.Bb3 Kh8 7.Bc4 Kh7
  =  (0.09)   Depth: 25   00:00:07  3718kN
1.h7+ Kh8 2.Kg5 Kg7 3.h8Q+ Kxh8 4.Kf6 Kg8 5.Ba2+ Kh7 6.Bb3 Kh8 7.Bc4 Kh7
  =  (0.09)   Depth: 26   00:00:08  4207kN
1.h7+ Kh8 2.Kg5 Kg7 3.h8Q+ Kxh8 4.Kf6 Kg8 5.Ba2+ Kh7 6.Bb3 Kh8 7.Bc4 Kh7
  =  (0.09)   Depth: 27   00:00:09  4771kN
1.h7+ Kh8 2.Kg5 Kg7 3.h8Q+ Kxh8 4.Kf6 Kg8 5.Ba2+ Kh7 6.Bb3 Kh8 7.Bc4 Kh7
  =  (0.09)   Depth: 28   00:00:10  5372kN
1.h7+ Kh8 2.Kg5 Kg7 3.h8Q+ Kxh8 4.Kf6 Kg8 5.Ba2+ Kh7 6.Bb3 Kh8 7.Bc4 Kh7
  =  (0.09)   Depth: 29   00:00:11  6006kN
1.h7+ Kh8 2.Kg5 Kg7 3.h8Q+ Kxh8 4.Kf6 Kg8 5.Ba2+ Kh7 6.Bb3 Kh8 7.Bc4 Kh7
  =  (0.09)   Depth: 30   00:00:12  6644kN
1.h7+ Kh8 2.Kg5 Kg7 3.h8Q+ Kxh8 4.Kf6 Kg8 5.Ba2+ Kh7 6.Bb3 Kh8 7.Bc4 Kh7
  =  (0.09)   Depth: 31   00:00:13  7495kN
1.h7+ Kh8 2.Kg5 Kg7 3.h8Q+ Kxh8 4.Kf6 Kg8 5.Ba2+ Kh7 6.Bb3 Kh8 7.Bc4 Kh7
  =  (0.09)   Depth: 32   00:00:14  8235kN
1.h7+ Kh8 2.Kg5 Kg7 3.h8Q+ Kxh8 4.Kf6 Kg8 5.Ba2+ Kh7 6.Bb3 Kh8 7.Bc4 Kh7
  =  (0.09)   Depth: 33   00:00:16  9017kN
1.h7+ Kh8 2.Kg5 Kg7 3.h8Q+ Kxh8 4.Kf6 Kg8 5.Ba2+ Kh7 6.Bb3 Kh8 7.Bc4 Kh7
  =  (0.09)   Depth: 34   00:00:17  9849kN
1.h7+ Kh8 2.Kg5 Kg7 3.h8Q+ Kxh8 4.Kf6 Kg8 5.Ba2+ Kh7 6.Bb3 Kh8 7.Bc4 Kh7
  =  (0.09)   Depth: 35   00:00:19  10743kN
1.h7+ Kh8 2.Kg5 Kg7 3.h8Q+ Kxh8 4.Kf6 Kg8 5.Ba2+ Kh7 6.Bb3 Kh8 7.Bc4 Kh7
  =  (0.09)   Depth: 36   00:00:20  11707kN
1.h7+ Kh8 2.Kg5 Kg7 3.h8Q+ Kxh8 4.Kf6 Kg8 5.Ba2+ Kh7 6.Bb3 Kh8 7.Bc4 Kh7
  =  (0.09)   Depth: 37   00:00:22  12740kN
1.h7+ Kh8 2.Kg5 Kg7 3.h8Q+ Kxh8 4.Kf6 Kg8 5.Ba2+ Kh7 6.Bb3 Kh8 7.Bc4 Kh7
  =  (0.09)   Depth: 38   00:00:24  13837kN
1.h7+ Kh8 2.Kg5 Kg7 3.h8Q+ Kxh8 4.Kf6 Kg8 5.Ba2+ Kh7 6.Bb3 Kh8 7.Bc4 Kh7
  =  (0.09)   Depth: 39   00:00:26  15036kN
1.h7+ Kh8 2.Kg5 Kg7 3.h8Q+ Kxh8 4.Kf6 Kg8 5.Ba2+ Kh7 6.Bb3 Kh8 7.Bc4 Kh7
  =  (0.09)   Depth: 40   00:00:28  16282kN
1.h7+ Kh8 2.Kg5 Kg7 3.h8Q+ Kxh8 4.Kf6 Kg8 5.Ba2+ Kh7 6.Bb3 Kh8 7.Bc4 Kh7
  =  (0.09)   Depth: 41   00:00:30  17608kN
1.h7+ Kh8 2.Kg5 Kg7 3.h8Q+ Kxh8 4.Kf6 Kg8 5.Ba2+ Kh7 6.Bb3 Kh8 7.Bc4 Kh7
  =  (0.09)   Depth: 42   00:00:32  19042kN
1.h7+ Kh8 2.Kg5 Kg7 3.h8Q+ Kxh8 4.Kf6 Kg8 5.Ba2+ Kh7 6.Bb3 Kh8 7.Bc4 Kh7
  =  (0.09)   Depth: 43   00:00:34  20551kN
1.h7+ Kh8 2.Kg5 Kg7 3.h8Q+ Kxh8 4.Kf6 Kg8 5.Ba2+ Kh7 6.Bb3 Kh8 7.Bc4 Kh7
  =  (0.09)   Depth: 44   00:00:37  22135kN
1.h7+ Kh8 2.Kg5 Kg7 3.h8Q+ Kxh8 4.Kf6 Kg8 5.Ba2+ Kh7 6.Bb3 Kh8 7.Bc4 Kh7
  =  (0.09)   Depth: 45   00:00:40  24051kN
1.h7+ Kh8 2.Kg5 Kg7 3.h8Q+ Kxh8 4.Kf6 Kg8 5.Ba2+ Kh7 6.Bb3 Kh8 7.Bc4 Kh7
  =  (0.09)   Depth: 46   00:00:42  25787kN
1.h7+ Kh8 2.Kg5 Kg7 3.h8Q+ Kxh8 4.Kf6 Kg8 5.Ba2+ Kh7 6.Bb3 Kh8 7.Bc4 Kh7
  =  (0.09)   Depth: 47   00:00:45  27884kN
1.h7+ Kh8 2.Kg5 Kg7 3.h8Q+ Kxh8 4.Kf6 Kg8 5.Ba2+ Kh7 6.Bb3 Kh8 7.Bc4 Kh7
  =  (0.09)   Depth: 48   00:00:48  29912kN
1.h7+ Kh8 2.Kg5 Kg7 3.h8Q+ Kxh8 4.Kf6 Kg8 5.Ba2+ Kh7 6.Bb3 Kh8 7.Bc4 Kh7
  =  (0.09)   Depth: 49   00:00:51  31977kN
1.h7+ Kh8 2.Kg5 Kg7 3.h8Q+ Kxh8 4.Kf6 Kg8 5.Ba2+ Kh7 6.Bb3 Kh8 7.Bc4 Kh7
  =  (0.09)   Depth: 50   00:00:54  34140kN
1.h7+ Kh8 2.Kg5 Kg7 3.h8Q+ Kxh8 4.Kf6 Kg8 5.Ba2+ Kh7 6.Bb3 Kh8 7.Bc4 Kh7
  =  (0.09)   Depth: 51   00:00:57  36554kN
1.h7+ Kh8 2.Kg5 Kg7 3.h8Q+ Kxh8 4.Kf6 Kg8 5.Ba2+ Kh7 6.Bb3 Kh8 7.Bc4 Kh7
  =  (0.09)   Depth: 52   00:01:00  39135kN
1.h7+ Kh8 2.Kg5 Kg7 3.h8Q+ Kxh8 4.Kf6 Kg8 5.Ba2+ Kh7 6.Bb3 Kh8 7.Bc4 Kh7
  =  (0.09)   Depth: 53   00:01:04  42170kN
1.h7+ Kh8 2.Kg5 Kg7 3.h8Q+ Kxh8 4.Kf6 Kg8 5.Ba2+ Kh7 6.Bb3 Kh8 7.Bc4 Kh7
  =  (0.09)   Depth: 54   00:01:09  46982kN
1.h7+ Kh8 2.Kg5 Kg7 3.h8Q+ Kxh8 4.Kf6 Kg8 5.Ba2+ Kh7 6.Bb3 Kh8 7.Bc4 Kh7
  =  (0.09)   Depth: 55   00:01:13  50495kN
1.h7+ Kh8 2.Kg5 Kg7 3.h8Q+ Kxh8 4.Kf6 Kg8 5.Ba2+ Kh7 6.Bb3 Kh8 7.Bc4 Kh7
  =  (0.09)   Depth: 56   00:01:18  55573kN
1.h7+ Kh8 2.Kg5 Kg7 3.h8Q+ Kxh8 4.Kf6 Kg8 5.Ba2+ Kh7 6.Bb3 Kh8 7.Bc4 Kh7
  =  (0.09)   Depth: 57   00:01:23  61709kN
1.h7+ Kh8 2.Kg5 Kg7 3.h8Q+ Kxh8 4.Kf6 Kg8 5.Ba2+ Kh7 6.Bb3 Kh8 7.Bc4 Kh7
  =  (0.09)   Depth: 58   00:01:31  70425kN
1.h7+ Kh8 2.Kg5 Kg7 3.h8Q+ Kxh8 4.Kf6 Kg8 5.Ba2+ Kh7 6.Bb3 Kh8 7.Bc4 Kh7
  =  (0.09)   Depth: 59   00:01:39  79706kN

(AMD, 15.02.2010)

Regards,
Gaмßito.
Parent - By yanquis1972 (****) [us] Date 2010-02-16 03:54
first of all, congratulations on the draw nevertheless. you did not know it but you exploited (fairly) a very well known & much complained about 'bug' - rybka doesn't understand wrong bishop endings. the reason for this is presumably because including the knowledge led to an overall decrease in engine elo.  if rybka was in the public eye & embarrassments such as this were witnessed enough (imagine such a thing happening in a big man-machine match), i am sure the appropriate change would be made and we would suffer an engine that is 3090 instead of 3100 elo or whathaveyou. the inconvenience of having to wait a few extra seconds for rybka to resolve a certain position every once in awhile might easily be superseded by the more accurate analysis & play for some, but vas apparently feels or felt otherwise.
Parent - By Felix Kling (Gold) [de] Date 2010-02-16 09:35
- - By domelius [de] Date 2010-02-16 04:00
Ok, Thanks. And success with the new versions!
Parent - - By BigBen (****) Date 2010-02-16 09:18 Edited 2010-02-16 09:30
Hi,
    My Rybka does not play 41.Bxf6? in fact no engine I tested did so I assume you had no TBases loaded .... Choices tend to be 41.Bf2 or 41.Bb6 and I think white will win

No other engine tested in fact played 41.Bxf6 with or without TBases loaded

Regards
Parent - - By yanquis1972 (****) [us] Date 2010-02-16 13:36
most users don't automatically have tablebases installed, certainly not 5 men. any engine should be smart enough to avoid this without tablebases & not need to use them as a crutch.
Parent - - By Mark (****) [us] Date 2010-02-16 14:33
I would think most users would have the full set of 5-man egtbs installed.  Nowadays, 7 GB is nothing...
Parent - - By yanquis1972 (****) [us] Date 2010-02-16 14:48
most heavy users yes, but probably not most users. it's a pain to DL 7GB of data from odd sites & no one wants to spend money on something that's free.  i think 5 man TBs are probably still relatively uncommon among the entire pool of people who use chess engines.
Parent - By Mark (****) [us] Date 2010-02-16 15:18
Maybe.  I just finished downloading the 6-man egtbs! 1.2TBs took quite awhile...
- - By domelius [de] Date 2010-02-16 09:39 Edited 2010-02-16 09:46
Hi Big Ben,

Ok, I dont have any tablebases. A tablebase for 7 pieces endgame, how big is that?

And a move from the bases is not an engine move at all.
Parent - By Felix Kling (Gold) [de] Date 2010-02-16 10:21
you just need the 5 men for that particular ending, as the engine uses the tablebases during calculation. you need just the files kbppk, that are rather small (50MB together). all 3-4-5 men tablebases are 7.1 GB total.

see http://kirill-kryukov.com/chess/tablebases-online/
Up Topic Rybka Support & Discussion / Rybka Discussion / Rybka 3 cannot find an obvious draw

Powered by mwForum 2.27.4 © 1999-2012 Markus Wichitill