Not logged inRybka Chess Community Forum
Up Topic The Rybka Lounge / Computer Chess / engine room and tablebases
- - By Hamlet (**) [fi] Date 2007-03-14 21:19
OK. For some reason I have finally visited in the notorious engine room.

I have played about 100 blitz games and lost very few games, but lately I have lost some very strange games.
For me it seems that there is somekind of problem with Fritz interface and tablebases. I have been in the 6-men or 5-men tablebases in the endgame. Sometimes situation has been draw and sometimes I have been winning. Here is one example (strange loss):

I lost in time in this situation where I was supposed to mate my opponent in 22 moves. For some reason my program stopped to make moves from tablebases.
(I tried to offer draw but no reply)

Is this knonw bug? and should I avoid using tablebases when playing in the engine room? Most of the players don't use tablebases and is this the reason?

Curious,

Hamlet

[Event "Rated game, 3m + 0s"]
[Site "Engine Room"]
[Result "0-1"]

{Rybka 2.3.1 mp 32-bit: 12.2 ply; 82kN/s                   Intel(R) Xeon(TM)
CPU 3.06GHz 3056MHz, (2 threads), RybkaII.ctg, pP\} 1. e4 {B/0 0} c5 {B/0 0} 2.
Nf3 {B/0 0} d6 {B/0 0} 3. d4 {B/0 0} cxd4 {B/0 0} 4. Nxd4 {B/0 0} Nf6 {B/0 0}
5. Nc3 {B/0 0} a6 {B/0 0} 6. f3 {B/0 0} e5 {B/0 0} 7. Nb3 {B/0 0} Be6 {B/0 0}
8. Be3 {B/0 0} Be7 {B/0 0} 9. Qd2 {B/0 0} O-O {B/0 0} 10. O-O-O {B/0 0} Nbd7 {
B/0 0} 11. g4 {B/0 0} b5 {B/0 0} 12. g5 {B/0 0} b4 {B/0 0} 13. Ne2 {B/0 0} Ne8
{B/0 0} 14. f4 {B/0 0} a5 {B/0 0} 15. f5 {B/0 0} Bc4 {B/0 0} 16. Ng3 {B/0 0}
Bxf1 {B/0 0} 17. Rhxf1 {B/0 0} a4 {B/0 0} 18. Na1 {B/0 0} a3 {B/0 0} 19. b3 {
B/0 0} Kh8 {B/0 0} 20. h4 {B/0 0} Rc8 {B/0 0} 21. f6 {B/0 0} gxf6 {B/0 0} 22.
Nf5 {B/0 0} Rc3 {B/0 0} 23. Kb1 {B/0 0} fxg5 {B/0 0} 24. hxg5 {B/0 0} Rxe3 {
B/0 0} 25. Qxe3 {B/0 0} Bxg5 {B/0 0} 26. Qe1 {B/0 0} Qa5 {B/0 0} 27. Nxd6 {
B/0 0} Nxd6 {B/0 0} 28. Rxd6 {B/0 0} Qb5 {B/0 0} 29. Rd5 {B/0 0} Qb7 {B/0 0}
30. c4 {B/0 0} bxc3 {0.32/12 4} 31. Nc2 {B/0 0} Qc6 {0.31/11 5} 32. Rf3 {
0.48/15 4} Bd2 {0.40/12 5} 33. Qf2 {0.69/16 1} Qe6 {0.47/11 2} 34. Nxa3 {
1.03/14 1} Qg4 {0.47/11 3} 35. Qe2 {1.03/14 1} Nf6 {0.80/11 7} 36. Rxe5 {
0.97/15 5} Ra8 {0.71/12 2} 37. Nc4 {1.57/15 10} Nh5 {1.43/10 18} 38. Rf2 {
1.52/18 10} Qg3 {(Qg4-g7) 1.31/11 4} 39. Qf3 {2.39/17 4} Qg1+ {2.20/13 1} 40.
Rf1 {2.46/18 2} Qd4 {2.20/13 0} 41. a3 {2.46/17 25} Nf4 {2.35/16 0} 42. Rg5 {
1.72/16 7} Rc8 {(f7-f6) 1.36/10 4} 43. Rh1 {2.02/11 1} Qd7 {(Rc8-c6) 1.49/10 1}
44. e5 {2.21/12 2} Rc6 {1.65/8 3} 45. Qe4 {2.54/15 2} Qd3+ {2.24/9 1} 46. Qxd3
{2.72/17 0} Nxd3 {2.22/12 0} 47. Rgg1 {2.79/16 0} Re6 {2.36/12 0} 48. Rh2 {
2.95/16 2} Nxe5 {(Re6-c6) 2.68/12 3} 49. Nxd2 {2.97/15 1} cxd2 {2.77/11 0} 50.
Rxd2 {3.08/15 0} h6 {(Re6-e8) 2.82/13 5} 51. Re2 {3.19/17 2} f6 {3.00/13 1} 52.
b4 {3.35/16 1} Rb6 {3.07/13 2} 53. Re4 {3.47/17 1} Kh7 {(Rb6-a6) 3.24/14 1} 54.
a4 {3.51/16 2} h5 {3.31/12 1} 55. Rh4 {3.75/17 0} Kh6 {3.37/10 1} 56. Rgh1 {
3.81/17 3} Kg6 {3.59/13 2} 57. b5 {3.90/16 3} f5 {(Rb6-b8) 3.70/13 3} 58. Rxh5
{4.11/15 2} Ng4 {(Ne5-f7) 3.81/13 1} 59. Kc2 {4.17/16 2} Nf6 {
(Rb6-b7) 3.94/12 2} 60. Rh8 {4.37/15 1} f4 {(Rb6-b7) 4.24/12 4} 61. Rb1 {
4.94/14 2} Rb7 {4.53/11 1} 62. b6 {5.14/14 0} f3 {5.08/12 3} 63. Rh3 {5.24/15 2
} f2 {5.17/11 0} 64. Rg3+ {5.17/14 2} Kf7 {5.24/12 1} 65. a5 {5.17/14 3} Ke6 {
(Rb7-e7) 5.19/10 3} 66. Rf3 {5.35/12 2} Nd5 {(Nf6-d7) 5.40/10 1} 67. Rxf2 {
5.86/11 0} Ke5 {5.98/10 2} 68. Kd3 {7.30/13 0} Ke6 {6.61/11 3} 69. Kc4 {
7.76/13 0} Rb8 {8.05/10 2} 70. b7 {13.33/12 1} Nc7 {10.09/12 2} 71. Kc5 {
18.28/13 0} Na6+ {16.49/14 9} 72. Kd4 {6.20/14 0} Rd8+ {7.25/10 2} 73. Kc3 {
7.15/14 3} Rd6 {(Rd8-b8) 7.47/12 0} 74. Ra1 {5.89/14 7} Ke7 {(Ke6-d7) 6.34/9 3}
75. Kc4 {6.42/12 1} Re6 {(Ke7-d7) 6.62/10 1} 76. Rg1 {6.77/11 1} Kd6 {
(Re6-d6) 6.93/9 2} 77. Rg7 {7.66/12 1} Rh6 {9.39/10 2} 78. Kb5 {7.67/13 2} Nc7+
{(Kd6-e5) 11.84/12 6} 79. Ka4 {6.54/13 1} Na6 {5.89/9 1} 80. Rc2 {6.85/13 1}
Kd5 {6.85/8 1} 81. Rd7+ {7.30/14 3} Ke4 {6.94/10 1} 82. Rd8 {7.43/14 1} Rh7 {
(Rh6-g6) 7.43/10 1} 83. Kb5 {7.35/12 0} Rxb7+ {7.66/10 1} 84. Kxa6 {#30/14 2}
Rg7 {7.84/11 1} 85. Re2+ {#29/11 0} Kf3 {7.66/11 1} 86. Re1 {6.96/12 1} Rg6+ {
(Rg7-h7) 7.28/9 2} 87. Kb5 {#32/11 1} Rg5+ {7.22/9 1} 88. Kb4 {7.82/13 1} Rg4+
{(Kf3-f2) 5.46/11 3} 89. Kb3 {7.91/13 1} Rg6 {7.63/11 1} 90. Rde8 {7.86/13 5}
Kg2 {7.24/9 1} 91. R1e6 {7.17/12 4} Rg3+ {(Rg6-g4) 7.85/9 2} 92. Kc4 {
14.43/12 3} Rg4+ {7.94/8 1} 93. Kb5 {7.25/12 1} Rg7 {(Rg4-g5+) 9.81/9 1} 94.
Rg6+ {#22/8 0} Rxg6 {Hamlet,Rybka 2.3.1 mp offers a draw Time  (Lag: Av=0.
51s, max=5.1s) 5.96/12 1} 0-1
Parent - - By SirN (*) [se] Date 2007-03-15 14:26 Edited 2007-03-15 14:44
Have you tested your tb's that they are ok? Do you have all 5p tb's? Do you have all the 6p endgames tb's that this endgame can transpose to? If no on any of the questions then it might be the reason for error!

Other than that I have no answer, but it has always been people claimning tb bug's, I have not experienced any problems and I'm just answerring to try to help you! What was the timecontrol, were you low on time? What GUI did you use(version date)? What engine did you use (why not full game header?)? I see after investigating the notsation that you had rybka 2.3.1 so answer is not needed

After looking at the game- I don't understand why white doesn't win..
Parent - - By Hamlet (**) [fi] Date 2007-03-15 15:11
Good questions. I have to check few things at home. Missing headers. I just don't want to disclose my playchess test handle.
Generally, I have used Wilhelm to check that tablebase files I have downloaded are not corrupted:
http://www.geocities.com/rba_schach2000/overview_english.htm

Regards,

Hamlet
Parent - - By Hamlet (**) [fi] Date 2007-03-15 18:16 Edited 2007-03-15 18:18
OK. I actually found four strange losses. Here they are:
(end positions)

I have lost 2 times with white. One is unneeded loss:

4R3/8/6r1/PK6/8/8/6k1/8 w - - 0 95


I have lost 9 times with black. Here are three uneeded:

8/5k1K/7P/8/8/7r/8/6R1 b - - 0 83


8/8/2k1Rp2/5r2/P7/4K3/8/8 b - - 0 92


8/6k1/8/7p/p7/5R2/3r4/5K2 b - - 0 54
Parent - - By Hamlet (**) [fi] Date 2007-03-15 18:37
And I have all 5-men tablebases.  Wilhelms says that my 5-men tablebase files are OK.

I also have krpkrp and krppkr and Wilhelm says that those are OK as well.

I don't have kqrkrp and kqrpkr, but two of losses was inside of 5-men files so that should not be the reason ?
and situation was not just before queening. I guess that when you are inside of tablebase files then Gui make moves and in these four situation gui stopped making moves. I was just watching how my time runs out.

Curious,

Hamlet
Parent - - By Hamlet (**) [fi] Date 2007-03-15 18:45
More info. Timecontrol was 3m+0s and I think that I had 10-20 sec time at least. I have fritz10+service pack.

I will reinstall fritz and service pack and let's see if it helps.

Regards,

Hamlet
Parent - - By SirN (*) [se] Date 2007-03-15 19:16
I remember now, if you increase the tb hash in Fritz GUI you get malfunction in old GUI (maybe still now also), the solution would be to use default 1MB and you'll have 100% function! Try it!
Parent - By Hamlet (**) [fi] Date 2007-03-16 00:00
OK. I have set it 256Mb. I will try. Thanks.

Regards,

hamlet
Up Topic The Rybka Lounge / Computer Chess / engine room and tablebases

Powered by mwForum 2.27.4 © 1999-2012 Markus Wichitill