Not logged inRybka Chess Community Forum
Up Topic Rybka Support & Discussion / Rybka Discussion / Rybka in Paderborn
1 2 3 4 5 Previous Next  
Parent - - By Venator (Silver) Date 2007-12-29 08:23
My feeling is that 27.axb3 is far too dangerous and probably already losing. Opening the a-file here
is suicide. No GM would consider that.

Regards, Jeroen
Parent - By M ANSARI (*****) Date 2007-12-29 11:08
Yes ofcourse ... also having connected passers is a long term plus.
Parent - - By Vasik Rajlich (Silver) Date 2007-12-30 12:52
Hi Jeroen,

I also would have preferred 22. Nf4 here, at the very least Rybka misevaluates the position. The newest Rybka handles this differently now.

Vas
Parent - - By Bouddha (****) Date 2007-12-30 13:35
Hello,

So what Rybka version did play in Paderborn ?
It was not the latest version ?

regards
Parent - - By Vasik Rajlich (Silver) Date 2007-12-30 13:59
The latest version played in Paderborn, but I continued working during the tournament.

Vas
Parent - - By Carl Bicknell (*****) Date 2007-12-30 14:37
It seems you have done a lot of work on the values of rook and minor piece - correct? Rybka seemed to sac the exchange every game
Parent - By Vasik Rajlich (Silver) Date 2008-01-03 14:38
Yes, Larry has been working on this.

Vas
Parent - - By Gaмßito (****) Date 2007-12-29 06:44
Yes, a short game against Jonny but with a fine book victory too.   
 
After all, Rybka did not have a problem to increase easily their advantage after Jonny took out of book to Rybka with 11. Qxa7.   

Regards,
Gaмßito.
Parent - - By Venator (Silver) Date 2007-12-29 11:46
A painful loss in round 5 against Toga Cluster. I didn't know Toga had a very good book, otherwise I wouldn't have
played this line. Everything was OK until Rybka went 30... Qh3? which looks like losing to me. When the black pawns
a6 and b5 are gone, nothing is going to stop a3-a4-a5 etc. Still, the weird ending with 3 pawns against rook might
have given some chances. Some people reported that Rybka could have played much better, so maybe somebody
can figure that out?

[Event "IPCCC Paderborn"]
[Site "Apeldoorn"]
[Date "2007.12.29"]
[Round "?"]
[White "Toga Cluster"]
[Black "Rybka"]
[Result "1-0"]
[ECO "B90"]
[PlyCount "150"]
[TimeControl "6300"]

{1MB, PaderbornB.ctg, Intel Core Duo T2300} 1. e4 {[%emt 0:00:00]} c5 {
[%emt 0:00:00]} 2. Nf3 {[%emt 0:00:13]} d6 {[%emt 0:00:00]} 3. d4 {
[%emt 0:00:13]} cxd4 {[%emt 0:00:00]} 4. Nxd4 {[%emt 0:00:15]} Nf6 {
[%emt 0:00:00]} 5. Nc3 {[%emt 0:00:16]} a6 {[%emt 0:00:00]} 6. f3 {
[%emt 0:00:26]} e5 {[%emt 0:00:00]} 7. Nb3 {[%emt 0:00:17]} Be6 {[%emt 0:00:00]
} 8. Be3 {[%emt 0:00:17]} Be7 {[%emt 0:00:00]} 9. Qd2 {[%emt 0:00:18]} O-O {
[%emt 0:00:00]} 10. O-O-O {[%emt 0:00:11]} Nbd7 {[%emt 0:00:00]} 11. g4 {
[%emt 0:00:21]} b5 {[%emt 0:00:00]} 12. g5 {[%emt 0:00:17]} Nh5 {[%emt 0:00:00]
} 13. Kb1 {[%emt 0:00:27]} Nb6 {[%emt 0:00:00]} 14. Nd5 {[%emt 0:00:21]} Bxd5 {
[%emt 0:00:00]} 15. exd5 {[%emt 0:00:16]} Qc7 {[%emt 0:00:00]} 16. Na5 {
[%emt 0:00:18]} Nxd5 {[%emt 0:00:00]} 17. Qxd5 {[%emt 0:00:20]} Qxa5 {
[%emt 0:00:00]} 18. Bd3 {[%emt 0:00:19]} Qd8 {[%emt 0:00:00]} 19. h4 {
[%emt 0:00:16]} Qd7 {[%eval 25,20] [%emt 0:01:43]} 20. Rhe1 {[%emt 0:00:09]} g6
{[%eval 24,23] [%emt 0:04:11]} 21. Be4 {[%emt 0:05:07]} Rad8 {
[%eval 24,21] [%emt 0:00:01]} 22. a3 {[%emt 0:02:44]} Qc8 {
[%eval 28,20] [%emt 0:01:33]} 23. Qd2 {[%emt 0:01:18]} Qc7 {
[%eval 22,20] [%emt 0:00:34]} 24. Qf2 {(Bd5) [%emt 0:04:06]} Rb8 {
[%eval 18,20] [%emt 0:04:56]} 25. c3 {(Bd5) [%emt 0:00:19]} f6 {
[%eval 11,18] [%emt 0:02:06]} 26. Rg1 {(Qc2) [%emt 0:03:50]} Kh8 {
[%eval 0,20] [%emt 0:01:44]} 27. Ba7 {(Qc2) [%emt 0:01:31]} Rbc8 {
[%eval 0,21] [%emt 0:01:23]} 28. Be3 {[%emt 0:02:23]} Rb8 {
[%eval 0,24] [%emt 0:00:01]} 29. Rd2 {(gxf6) [%emt 0:02:51]} Qd7 {
[%eval 0,19] [%emt 0:01:40]} 30. Rgd1 {[%emt 0:01:38]} Qh3 {
[%eval 11,21] [%emt 0:03:49]} 31. Ba7 {[%emt 0:00:47]} Rbd8 {
[%eval 18,22] [%emt 0:06:26]} 32. Qb6 {[%emt 0:00:17]} Qxh4 {
[%eval 24,22] [%emt 0:02:19]} 33. Qxa6 {[%emt 0:02:23]} Ng3 {
[%eval 37,23] [%emt 0:08:13]} 34. Bf2 {[%emt 0:03:15]} Qxg5 {
[%eval 37,23] [%emt 0:00:05]} 35. Bc6 {(Bd5) [%emt 0:03:19]} f5 {
[%eval 30,20] [%emt 0:03:08]} 36. Bb6 {[%emt 0:02:55]} Rb8 {
[%eval 54,22] [%emt 0:03:42]} 37. Bc7 {[%emt 0:03:14]} Qe3 {
[%eval 72,21] [%emt 0:01:18]} 38. Rd3 {(Bxb8) [%emt 0:02:05]} Qf2 {
[%eval 78,21] [%emt 0:01:21]} 39. Bxb8 {[%emt 0:02:11]} Rxb8 {
[%eval 120,22] [%emt 0:03:51]} 40. Bxb5 {[%emt 0:01:29]} Ne2 {
[%eval 119,21] [%emt 0:04:18]} 41. a4 {(Qc6) [%emt 0:03:02]} Qc5 {
[%eval 165,21] [%emt 0:02:09]} 42. Rd5 {[%emt 0:00:31]} Nxc3+ {
[%eval 168,21] [%emt 0:00:59]} 43. bxc3 {[%emt 0:01:15]} Qa3 {
[%eval 168,19] [%emt 0:00:23]} 44. Qc6 {[%emt 0:04:24]} Qxa4 {
[%eval 195,22] [%emt 0:00:01]} 45. R1d2 {[%emt 0:02:36]} Qa7 {
[%eval 204,22] [%emt 0:00:39]} 46. Ra2 {[%emt 0:02:00]} Qb7 {
[%eval 206,24] [%emt 0:00:25]} 47. Kc2 {[%emt 0:01:48]} Kg7 {
[%eval 207,24] [%emt 0:00:01]} 48. c4 {[%emt 0:01:28]} Kh6 {
[%eval 207,23] [%emt 0:00:01]} 49. Qxb7 {[%emt 0:01:57]} Rxb7 {
[%eval 223,25] [%emt 0:00:01]} 50. Ra6 {(Bc6) [%emt 0:01:18]} Kg5 {
[%eval 223,22] [%emt 0:01:32]} 51. Bc6 {[%emt 0:00:38]} Rc7 {
[%eval 223,23] [%emt 0:02:55]} 52. c5 {[%emt 0:00:13]} dxc5 {
[%eval 248,24] [%emt 0:01:08]} 53. Rxe5 {[%emt 0:00:41]} Bf8 {
[%eval 248,24] [%emt 0:00:08]} 54. Kd3 {[%emt 0:01:26]} Kf4 {
[%eval 242,24] [%emt 0:00:01]} 55. Rd5 {[%emt 0:01:14]} Kg3 {
[%eval 242,24] [%emt 0:00:00]} 56. Kc4 {[%emt 0:02:01]} Bh6 {
[%eval 255,23] [%emt 0:00:27]} 57. Kxc5 {(Re5) [%emt 0:01:21]} Be3+ {
[%eval 264,23] [%emt 0:03:08]} 58. Kb5 {(Kc4) [%emt 0:00:14]} Re7 {
[%eval 234,23] [%emt 0:01:32]} 59. Ba8 {(Kc4) [%emt 0:00:15]} g5 {
[%eval 216,23] [%emt 0:01:21]} 60. Rxf5 {[%emt 0:01:59]} h5 {
[%eval 224,23] [%emt 0:00:01]} 61. Ra4 {[%emt 0:01:04]} Re8 {
[%eval 240,23] [%emt 0:01:32]} 62. Bd5 {[%emt 0:00:15]} Rb8+ {
[%eval 240,21] [%emt 0:01:40]} 63. Kc4 {(Kc6) [%emt 0:00:22]} Rb2 {
[%eval 297,21] [%emt 0:01:06]} 64. Kc3 {(Kd3) [%emt 0:01:16]} Rb5 {
[%eval 417,23] [%emt 0:02:12]} 65. Kd3 {[%emt 0:00:23]} Bf4 {
[%eval 449,25] [%emt 0:01:38]} 66. Raxf4 {[%emt 0:00:14]} gxf4 {
[%eval 502,26] [%emt 0:01:37]} 67. Rxh5 {[%emt 0:00:13]} Rb8 {
[%eval 502,25] [%emt 0:00:00]} 68. Ke4 {(Rg5+) [%emt 0:01:39]} Rb4+ {
[%eval 509,25] [%emt 0:00:27]} 69. Kf5 {[%emt 0:00:48]} Kf2 {
[%eval 509,26] [%emt 0:00:01]} 70. Kg4 {(Ke5) [%emt 0:00:00]} Ke3 {
[%eval 512,21] [%emt 0:01:11]} 71. Re5+ {[%emt 0:00:41]} Kd2 {
[%eval 512,19] [%emt 0:00:00]} 72. Be4 {(Re7) [%emt 0:01:15]} Ke3 {
[%eval 512,22] [%emt 0:00:37]} 73. Rd5 {[%emt 0:00:45]} Rb8 {
[%eval 512,22] [%emt 0:00:46]} 74. Rd3+ {[%emt 0:00:38]} Kf2 {
[%eval 512,20] [%emt 0:00:49]} 75. Kxf4 {[%emt 0:00:27]} Rf8+ {
[%eval 512,19] [%emt 0:00:41]} 1-0
Parent - By richbell (**) Date 2007-12-29 12:05
Parent - By BB (****) Date 2007-12-29 12:52 Edited 2007-12-29 13:09
Hiarcs (with white) beat Shredder in Round 5, leaving it and Cluster Toga with Rybka at 3.5/5, Jonny and Spike at 3/5. Round 6 Pairings: Rybka-Isichess, Spike-Cluster Toga, Hiarcs-Gandalf. Hiarcs already played Spike, so Rybka-Spike is likely for Round 7, and perhaps Jonny-Hiarcs. Of course, this could change if Spike loses and Diep beats Jonny. In any case, Cluster Toga will now have played everyone except the bottom three.
Parent - - By Venator (Silver) Date 2007-12-29 14:44
In round 4 Rybka beat Isichess with white. Here is the game:

[Event "IPCCC Paderborn"]
[Site "Apeldoorn"]
[Date "2007.12.29"]
[Round "?"]
[White "Rybka"]
[Black "Isichess"]
[Result "1-0"]
[ECO "C88"]
[PlyCount "95"]
[TimeControl "6300"]

{1MB, PaderbornB.ctg, Intel Core Duo T2300} 1. e4 {[%emt 0:00:00]} e5 {
[%emt 0:00:17]} 2. Nf3 {[%emt 0:00:00]} Nc6 {[%emt 0:00:08]} 3. Bb5 {
[%emt 0:00:00]} a6 {[%emt 0:00:11]} 4. Ba4 {[%emt 0:00:00]} Nf6 {[%emt 0:00:18]
} 5. O-O {[%emt 0:00:00]} Be7 {[%emt 0:00:11]} 6. Re1 {[%emt 0:00:00]} b5 {
[%emt 0:00:09]} 7. Bb3 {[%emt 0:00:00]} O-O {[%emt 0:00:08]} 8. h3 {
[%emt 0:00:00]} Bb7 {[%emt 0:00:20]} 9. d3 {[%emt 0:00:00]} d6 {[%emt 0:00:09]}
10. a3 {[%emt 0:00:00]} Nd4 {[%emt 0:00:12]} 11. Nxd4 {[%emt 0:00:00]} exd4 {
[%emt 0:00:08]} 12. c3 {[%eval 15,20] [%emt 0:02:38]} dxc3 {(c5) [%emt 0:00:13]
} 13. Nxc3 {[%eval 24,21] [%emt 0:04:17]} c5 {(Re8) [%emt 0:00:10]} 14. a4 {
[%eval 21,20] [%emt 0:06:36]} b4 {(Qd7) [%emt 0:16:31]} 15. Ne2 {
[%eval 26,20] [%emt 0:02:05]} d5 {(Nd7) [%emt 0:07:04]} 16. e5 {
[%eval 46,21] [%emt 0:01:43]} Ne8 {[%emt 0:02:40]} 17. Be3 {
[%eval 39,22] [%emt 0:00:58]} f6 {(Nc7) [%emt 0:04:36]} 18. Qc2 {
[%eval 44,19] [%emt 0:03:35]} fxe5 {[%emt 0:04:13]} 19. Bxc5 {
[%eval 42,22] [%emt 0:00:01]} Bd6 {(Rc8) [%emt 0:02:44]} 20. Bxd6 {
[%eval 73,20] [%emt 0:01:49]} Qxd6 {[%emt 0:01:46]} 21. Ng3 {
[%eval 76,20] [%emt 0:00:05]} Rf7 {[%emt 0:00:58]} 22. Qe2 {
[%eval 88,22] [%emt 0:00:18]} Qf6 {[%emt 0:00:56]} 23. Qxe5 {
[%eval 104,21] [%emt 0:02:40]} Qxf2+ {[%emt 0:00:28]} 24. Kh2 {
[%eval 104,20] [%emt 0:00:21]} Nf6 {(Qf6) [%emt 0:01:21]} 25. Ne4 {
[%eval 131,19] [%emt 0:00:38]} Qh4 {(Nxe4) [%emt 0:00:20]} 26. Qe6 {
[%eval 148,20] [%emt 0:00:45]} Rd8 {(Nxe4) [%emt 0:02:34]} 27. Rf1 {
[%eval 156,21] [%emt 0:01:03]} Bc8 {[%emt 0:02:31]} 28. Qb6 {
[%eval 153,22] [%emt 0:00:49]} Rff8 {[%emt 0:01:22]} 29. Nxf6+ {
[%eval 156,21] [%emt 0:00:20]} gxf6 {[%emt 0:01:17]} 30. Rae1 {
[%eval 156,22] [%emt 0:01:00]} Kh8 {(f5) [%emt 0:02:10]} 31. Re7 {
[%eval 174,20] [%emt 0:01:22]} Qg5 {[%emt 0:02:34]} 32. Qa7 {
[%eval 187,21] [%emt 0:00:01]} Bf5 {(Qh6) [%emt 0:01:39]} 33. Rf3 {
[%eval 200,21] [%emt 0:00:55]} h5 {[%emt 0:00:27]} 34. Qd4 {
[%eval 207,20] [%emt 0:00:55]} Bh7 {(Rd6) [%emt 0:01:52]} 35. Rg3 {
[%eval 256,21] [%emt 0:00:44]} Qh6 {[%emt 0:01:04]} 36. Ra7 {
[%eval 260,21] [%emt 0:00:01]} Rg8 {(Ra8) [%emt 0:01:14]} 37. Rxg8+ {
[%eval 294,19] [%emt 0:00:34]} Bxg8 {[%emt 0:01:22]} 38. Rxa6 {
[%eval 328,21] [%emt 0:00:05]} Kg7 {(Rf8) [%emt 0:04:44]} 39. a5 {
[%eval 321,20] [%emt 0:00:52]} Qg5 {[%emt 0:01:18]} 40. Rb6 {
[%eval 328,22] [%emt 0:00:01]} Ra8 {(h4) [%emt 0:01:39]} 41. a6 {
[%eval 362,18] [%emt 0:00:24]} h4 {(Qe5+) [%emt 0:01:59]} 42. Qxb4 {
[%eval 386,18] [%emt 0:00:24]} Qg3+ {[%emt 0:01:00]} 43. Kh1 {
[%eval 423,20] [%emt 0:00:00]} Re8 {(Bf7) [%emt 0:01:26]} 44. Qg4+ {
[%eval 508,18] [%emt 0:00:13]} Qxg4 {[%emt 0:01:15]} 45. hxg4 {
[%eval 512,21] [%emt 0:00:00]} Re1+ {[%emt 0:02:56]} 46. Kh2 {
[%eval 0,4] [%emt 0:00:00]} Ra1 {[%emt 0:01:11]} 47. Rb7+ {
[%eval 514,19] [%emt 0:00:00]} Kh6 {(Kg6) [%emt 0:01:06]} 48. a7 {
[%eval 526,14] [%emt 0:00:13]} 1-0
Parent - - By BB (****) Date 2007-12-29 14:58

> 1. e4 e5 2. Nf3 Nc6 3. Bb5 a6 4. Ba4 Nf6 5. O-O Be7 6. Re1 b5 7. Bb3 O-O 8. h3 Bb7 9. d3 d6 10. a3 Nd4


11. Nxd4 {[%emt 0:00:00]} exd4 {[%emt 0:00:08]} 12. c3 {[%eval 15,20] [%emt 0:02:38]} dxc3 {(c5) [%emt 0:00:13]}
If I read this correctly, Nxd4 was in book, but the position after the (forced) response exd4 was not? Doesn't make a lot of sense to me, but then, I don't make books (yet).
Parent - By Venator (Silver) Date 2007-12-29 15:06
It is not always necessary to have a lot of book moves. In this case I decided 'let Rybka play on her own'.
If she likes the white position and can find good moves herself, just let her play.
Parent - - By Venator (Silver) Date 2007-12-29 15:02
Spike-Cluster Toga will be a draw, that leaves the following standings before the last round:

1. Rybka, Hiarcs  4,5
3. Cluster Toga, Jonny  4
5. Shredder, Spike  3,5

Pairing Rybka vs Spike looks logical for the last round. Hiarcs will probably meet either Jonny
or Diep. Tournament is decided on Buchholz and this is favourable for Rybka, as Hiarcs has
won against ParSOS that has only 0,5 points.

See you tomorrow.
Parent - - By Venator (Silver) Date 2007-12-29 15:04
But.... If ParSOS beats Isichess, that advantage is gone of course :-)
Parent - By richbell (**) Date 2007-12-29 15:12
Could someone tell me what's "fine score" pls?
Parent - - By Venator (Silver) Date 2007-12-29 16:21
Last round pairings:

Jonny - Hiarcs
Diep - Rybka

Apparently a pairing Spike - Rybka wasn't possible. I am not too pleased with this, as Jonny has 4 points
and Diep only 2,5. That is a bad pairing from the Buchholz point of view.
Parent - - By Harvey Williamson (*****) Date 2007-12-29 16:25
Good Moaning Jeroen,

I am also surprised by this. Do you know why the pairing with spike was not possible?

Good Luck in final round,

Harvey
Parent - - By Felix Kling (Gold) Date 2007-12-29 16:36
maybe sth. with the colours...
Parent - - By Harvey Williamson (*****) Date 2007-12-29 17:47
That is my guess :)
Parent - By Felix Kling (Gold) Date 2007-12-29 18:03
I guess there's no other possible reason :)
Parent - - By Vempele (Silver) Date 2007-12-29 18:05
In most pairing systems, colours don't matter in the final round.

In any case, it's a huge mistake from the organizers. They should have made the pairings by hand; machine pairings (using the normal software anyway) just aren't reliable when you have this few participants and too many rounds...
Parent - - By Felix Kling (Gold) Date 2007-12-29 18:25
"In most pairing systems, colours don't matter in the final round."
At least in swiss system they may be important afaik... E.g. 2 players could have 2 games with white too much or the last two games could be with the same colour - or does this rule not apply in the last round?!
Parent - By Vempele (Silver) Date 2007-12-29 18:29 Edited 2007-12-29 18:34

> - or does this rule not apply in the last round?!


FIDE Handbook:

B.2  
No player's colour difference will become >+2 or <-2.
No player will receive the same colour three times in row.
...
Note: B2, B5 and B6 do not apply when pairing players with a score of over 50% in the last round.

That last bit about 50% could well be the problem, though.
Parent - By BB (****) Date 2007-12-30 08:44
maybe sth. with the colours...

This should not be the case. If you look at the round 6 pairings, they seem to have been made (deliberately?) to ensure that each engine had 3 whites and blacks after 6 rounds. Rybka and Spike were both BWBWBW in the first six rounds.
Parent - By BB (****) Date 2007-12-30 08:57
Do you know why the pairing with spike was not possible?

I should have expected to find the answer at the Hiarcs forum. :)

Here is a description: pairing Rybka-Spike and Jonny-Hiarcs forces Diep-Shredder, as Diep's only other non-opponents are Rybka and Hiarcs - but now Cluster Toga and the bottom three are left to pair, and Gandalf, Isichess, and ParsOS have already all played each other. So the inability to pair the bottom (without repeats) forces reallocation at the top.

Alternatively, Rybka-Spike, Diep-Hiarcs, Jonny-Isichess, Shredder-Gandalf, Cluster Toga-ParsOS would have worked, so it's a "decision" on whether to have Rybka-Spike or Jonny-Hiarcs --- but you can't have them both. [This being said, Jonny has 4 points compared to 3.5 for Spike, so it makes sense for it to play one of the competitors on 4.5 points].

I hate the Swiss system, particularly when the number of rounds is more than half than the number of players.
Parent - By Roland Rösler (****) Date 2007-12-29 21:31
Keep cool, Jeroen. I see perfect pairings. There is no way for Hiarcs to win against Jonny. So win against Diep and you have a nice day.
Parent - - By Ernst (***) Date 2007-12-29 23:31
Pairings weren't particularly lucky for Shredder either, playing black against Rybka, ClusterToga and Hiarcs. Those things happen and better luck next time.
Parent - By NATIONAL12 (Gold) Date 2007-12-29 23:40
at least shredder, hiarcs and rybka took the risk knowing this could happen, but i cannot understand why rybka did not try to get toga out of book earlier,like harvey came up with against rybka. because a prog had a crap book a few months ago does not mean it still has.
Parent - - By Harvey Williamson (*****) Date 2007-12-30 07:26 Edited 2007-12-30 07:30
Hi Jeroen,

I hope you are not all to hungover after the group outing last night. I hear there is a new rule for deciding the winner if we are = on points at the end. The team with the slower hardware gets 1st prize.

Happy New Year,

Harvey
Parent - - By turbojuice1122 (Gold) Date 2007-12-30 07:58
It's not the team with the better recipes?
Parent - - By Harvey Williamson (*****) Date 2007-12-30 07:59
:) That would work!
Parent - By Venator (Silver) Date 2007-12-30 08:44
Then I suggest that the program that has been used in most recipes should win :-)
Three Michelin stars for Rybka.....
Parent - By BB (****) Date 2007-12-30 09:26
Diep - Rybka
1. d4 d5 2. c4 c6 3. Nf3 Nf6 4. Nc3 e6 5. e3 Nbd7 6. Bd3 dxc4 7. Bxc4 b5 8. Bd3 Bb7 9. O-O a6 10. e4 c5 11. d5 Qc7 12. dxe6 fxe6 13. Bc2 c4 14. Ng5 Nc5 15. Qf3 h6 16. Qh3 O-O-O 17. e5 hxg5 18. Qxh8 Qxe5 19. Qh3 g4 20. Qg3 Qh5 21. Bf4 Nd3

Looks like a long book line with an exchange sac - Rybka/Shredder/Zappa evals are diverging a lot here, but Black appears to have the edge.
Parent - - By FWCC (***) Date 2007-12-29 15:09 Edited 2007-12-29 15:24
A painful and mysterious loss.Perhaps Rybka could have played more agressively on the Queenside earlier and also with 25...a5 26.Ba7 b4 27.Bxb8 Rxb8 28.cxb4 axb4 29.Rc1 Qa5  with attack(can't post position with this weak comp).Toga overpowered at the end with the bishops.37...Qe3 seems to have been a lemon,maybe 37...Nf1 was better though white is still on top.Also 44...Qb3+ was a better try as 44...Qxa4 simply losses outright.Better luck next round!! Rybka should have her way with Isichess next round,you can count on that as a Rybka win.
Parent - - By richbell (**) Date 2007-12-29 20:32
Could someone tell me what the 'Fine' score is?? mentioned in Paderborn scoresheet?
Parent - - By Bothkill (*) Date 2007-12-29 21:01
see: http://en.wikipedia.org/wiki/Buchholz_chess_rating

"It sums up the score of the players' opponents and thus favors those who have confronted better opponents."
Parent - By richbell (**) Date 2007-12-29 23:23
Thanks Bothkill, got it.
Parent - - By Uri Blass (*****) Date 2007-12-29 22:00
I do not know but I dislike 37...Qe3

37...Rbc8 seems to me better and commercial rybka also likes this move but maybe I did not give her enough time(45 minutes on my slow hardware may be the same as 3 minutes with fast hardware and rybka used in the game 4 minutes and 32 seconds if I include pondering
and this is without mentioning hash from previous search and bigger hash that rybka probably used relative to the 512 mbytes hash that I gave rybka).

Rybka 2.3.2a 32-bit, 80'/40+40'/20+20'
1r3r1k/2B1b2p/Q1Bp2p1/1p2ppq1/8/P1P2Pn1/1P1R4/1K1R4 b - - 0 1


Analysis by Rybka 2.3.2a 32-bit :

37...Qg5-e3 38.Bc7xb8
  =  (0.17)   Depth: 5   00:00:00
37...Ng3-f1 38.Rd2-d3
  =  (0.14)   Depth: 5   00:00:00
37...Ng3-f1 38.Rd2-d3 Rb8-c8
  ²  (0.27)   Depth: 6   00:00:00  7kN
37...Ng3-f1 38.Rd2xd6 Nf1-e3 39.Bc7xb8 Rf8xb8 40.Qa6-a7
  ²  (0.29)   Depth: 7   00:00:00  36kN
37...Ng3-f1 38.Rd2xd6 Nf1-e3 39.Bc7xb8 Rf8xb8 40.Qa6-a7 Be7xd6
  ²  (0.40)   Depth: 8   00:00:00  43kN
37...Qg5-e3 38.Bc7xb8 Rf8xb8 39.Rd2-d5 Kh8-g7 40.Rd5xb5 Qe3-e2
  ²  (0.38)   Depth: 8   00:00:00  50kN
37...Qg5-e3 38.Bc7xb8 Rf8xb8 39.Rd2-d5 Kh8-g7 40.Rd5xb5 Qe3-e2 41.Rd1-d2
  ²  (0.56)   Depth: 9   00:00:01  80kN
37...Ng3-f1 38.Rd2xd6 Nf1-e3 39.Bc7xb8 Rf8xb8 40.Qa6-a7 Be7xd6 41.Rd1xd6 Rb8-c8
  ²  (0.43)   Depth: 9   00:00:01  91kN
37...Rb8-c8 38.Bc7xd6 Be7xd6 39.Rd2xd6 Rc8-b8 40.Qa6-a7 Qg5-h4 41.Qa7-c5
  ²  (0.38)   Depth: 9   00:00:02  116kN
37...Rb8-c8 38.Bc7xd6 Be7xd6 39.Rd2xd6 Rc8-b8 40.Qa6-a7 Qg5-h4 41.Qa7-c5 Qh4-e7
  ²  (0.38)   Depth: 10   00:00:02  147kN
37...Rb8-c8 38.Bc7xd6 Be7xd6 39.Rd2xd6 Rc8-b8 40.Qa6-a7 Qg5-h4 41.Qa7-c5 Qh4-e7 42.Bc6xb5 e5-e4
  ²  (0.39)   Depth: 11   00:00:03  211kN
37...Rb8-c8 38.Bc7xd6 Be7xd6 39.Rd2xd6 Rc8-b8 40.Qa6-a7 Qg5-h4 41.Qa7-e3 Qh4-e7 42.Rd6-d7 Qe7-f6
  ²  (0.52)   Depth: 12   00:00:08  422kN
37...Rb8-c8 38.Bc7xd6 Be7xd6 39.Rd2xd6 Rc8-b8 40.Qa6-a7 Qg5-h4 41.Qa7-e3 Qh4-f4 42.Qe3-c5 Kh8-g7 43.Rd1-e1
  ²  (0.49)   Depth: 13   00:00:13  667kN
37...Rb8-c8 38.Bc7xd6 Be7xd6 39.Rd2xd6 Rc8-b8 40.Qa6-a7 Qg5-h4 41.Qa7-e3 Qh4-f4 42.Qe3-c5 Kh8-g7 43.Rd1-e1 h7-h5
  ²  (0.58)   Depth: 14   00:00:23  1168kN
37...Qg5-e3 38.Bc7xb8 Rf8xb8 39.Rd2-d5 b5-b4 40.a3xb4 Kh8-g7 41.Rd5-a5 h7-h5 42.Qa6-a7 Qe3xa7 43.Ra5xa7 Kg7-f6 44.b4-b5
  ²  (0.52)   Depth: 14   00:00:34  1747kN
37...Qg5-e3 38.Bc7xb8 Rf8xb8 39.Rd2-d5 b5-b4 40.a3xb4 Kh8-g7 41.Rd5-a5 h7-h5 42.Qa6-c4 Qe3-e2 43.Qc4xe2 Ng3xe2 44.Rd1-h1
  ²  (0.56)   Depth: 15   00:00:45  2264kN
37...Qg5-e3 38.Bc7xb8 Rf8xb8 39.Rd2-d5 Kh8-g7 40.Qa6-a5 Be7-d8 41.Qa5-b4 Bd8-e7 42.Rd5xb5 Qe3-e2 43.Rd1-g1 Rb8xb5 44.Qb4xb5
  ²  (0.56)   Depth: 16   00:01:39  4772kN
37...Qg5-e3 38.Bc7xb8 Rf8xb8 39.Rd2-d5 Kh8-g7 40.Qa6-a5 Be7-d8 41.Qa5-b4 Bd8-e7 42.Rd5xb5 Qe3-e2 43.Rd1-g1 Rb8xb5 44.Qb4xb5
  ²  (0.52)   Depth: 17   00:02:56  9042kN
37...Qg5-e3 38.Rd2-d3 Qe3-f2 39.Bc7xb8 Rf8xb8 40.Bc6xb5 Qf2-c5 41.Rd3-d5 Qc5-c7 42.Qa6-c6 Qc7xc6 43.Bb5xc6 Kh8-g7 44.Rd5-b5
  ±  (0.84)   Depth: 18   00:06:05  17352kN
37...Rb8-c8 38.Bc7xd6 Be7xd6 39.Rd2xd6 Rc8-b8 40.Qa6-a7 Qg5-h4 41.Qa7-e3 Qh4-f4 42.Qe3-c5 Kh8-g7 43.Rd1-e1 e5-e4 44.Rd6-d7+
  ±  (0.82)   Depth: 18   00:12:56  41139kN
37...Rb8-c8 38.Bc7xd6 Be7xd6 39.Rd2xd6 Rc8-b8 40.Qa6-a7 Qg5-h4 41.Qa7-e3 Qh4-f4 42.Qe3-c5 Kh8-g7 43.Rd1-e1 e5-e4 44.Rd6-d7+
  ±  (0.79)   Depth: 19   00:20:29  64331kN
37...Rb8-c8 38.Bc7xd6 Be7xd6 39.Rd2xd6 e5-e4 40.Qa6xb5 Qg5-f4 41.Qb5-b4 Kh8-g7 42.f3xe4 f5xe4 43.Rd6-d7+ Kg7-h6 44.Qb4-b7
  ±  (0.84)   Depth: 20   00:37:46  119093kN

(,  29.12.2007)

Parent - By Ernst (***) Date 2007-12-29 23:29
Zappa Mexico would have played the same Qe3. Hardware used Q6600@3GHz, 2GB hash.

Analysis by Zappa Mexico X64:

1...Nf1 2.Rd3 Rbc8 3.Bxd6 Bxd6 4.Rxf1 Rcd8 5.Rfd1
  =  (0.00)   Depth: 9/26   00:00:03  562kN
1...Nf1 2.Rd3 Rbc8 3.Bxd6 Bxd6 4.Rxf1 Rcd8
  =  (0.00)   Depth: 9/33   00:00:03  758kN
1...Nf1 2.Rd3 Rbc8 3.Bxd6 Bxd6 4.Rxd6 Ne3 5.Qxb5 Qf4 6.Rh1
  +/=  (0.31)   Depth: 10/33   00:00:03  1231kN
1...Qe3 2.Bxb8 Rxb8 3.Rd5 Bf8 4.R1d3 Qg1+ 5.Ka2 Ne2
  =  (0.21)   Depth: 10/33   00:00:03  2741kN
1...Qe3 2.Bxb8 Rxb8 3.Rd5 Bf8 4.R1d3 Qg1+ 5.Ka2 Ne2
  =  (0.21)   Depth: 10/33   00:00:04  3112kN
1...Qe3 2.Bxb8 Rxb8 3.Rd5 Bf8 4.Rxb5 Qe2 5.Rd2 Qf1+ 6.Kc2 e4 7.fxe4 Nxe4
  +/=  (0.48)   Depth: 11/33   00:00:04  4619kN
1...Qe3 2.Bxb8 Rxb8 3.Rd5 Bf8 4.Rxb5 Qe2 5.Rd2 Qf1+ 6.Kc2 e4 7.fxe4 Nxe4
  +/=  (0.48)   Depth: 11/37   00:00:06  8582kN
1...Qe3 2.Bxb5 Ra8 3.Qb7 Rfc8 4.Bxd6 Bxd6 5.Rxd6 Rcb8 6.Qd5
  +/=  (0.60)   Depth: 12/39   00:00:09  18750kN
1...Qe3 2.Bxb5 Ra8 3.Qb7 Rfc8 4.Bxd6 Bxd6 5.Rxd6 Rcb8 6.Qd5
  +/=  (0.60)   Depth: 12/39   00:00:12  25217kN
1...Qe3 2.Bxb5 Ra8 3.Qb7 Rfc8 4.Bxd6 Bxd6 5.Rxd6 Rcb8 6.Qd5
  +/=  (0.60)   Depth: 13/39   00:00:16  35846kN
1...Qe3 2.Bxb5 Ra8 3.Qb7 Rfc8 4.Bxd6 Bxd6 5.Rxd6 Rcb8 6.Qd5
  +/=  (0.60)   Depth: 13/39   00:00:19  46431kN
1...Qe3 2.Bxb5 Qc5 3.Bxb8 Rxb8 4.Rd5 Qc7 5.Qc6 Qxc6 6.Bxc6 h5 7.Rb5 Rxb5 8.Bxb5
  +/-  (0.92)   Depth: 14/48   00:00:38  97152kN
1...f4 2.Bxb8 Rxb8 3.Rd5 Qf5+ 4.Ka1 Rf8 5.Bxb5 Rb8 6.Qc6 Qf8 7.R5d2 Nf5
  +/-  (0.75)   Depth: 14/48   00:00:47  126mN
1...f4 2.Bxb8 Rxb8 3.Rd5 Qf5+ 4.Ka1 Rf8 5.Bxb5 Rb8 6.Qc6 Qf8 7.R5d2 Nf5
  +/-  (0.75)   Depth: 14/48   00:01:02  170mN
1...f4 2.Rd5 Qf5+ 3.Ka1 Rbc8 4.Bxd6 Bxd6 5.Rxd6 Qh5 6.Rd7 e4 7.Qb6 Qh4 8.fxe4 f3 9.R1d6 Rxc6
  +/-  (0.83)   Depth: 15/50   00:01:23  231mN
1...Qe3 2.Bxb8 Rxb8 3.Bxb5 Qc5 4.Rd5 Qc7 5.Ba4 Qxc3 6.Rb5 Rxb5 7.Qxb5 Qc8 8.Rc1
  +/-  (0.80)   Depth: 15/50   00:02:06  359mN
1...Qe3 2.Bxb8 Rxb8 3.Bxb5 Qc5 4.Rd5 Qc7 5.Ba4 Qxc3 6.Rb5 Rxb5 7.Qxb5 Qc8 8.Rc1
  +/-  (0.80)   Depth: 15/50   00:02:18  392mN
1...Qe3 2.Bxb8 Rxb8 3.Bxb5 Qc5 4.Rd5 Qc7 5.Qc6 Qxc6 6.Bxc6 h5 7.Kc2 h4 8.b4 Kg7
  +/-  (0.77)   Depth: 16/50   00:02:58  508mN
1...Qe3 2.Bxb8 Rxb8 3.Bxb5 Qc5 4.Rd5 Qc7 5.Qc6 Qxc6 6.Bxc6 h5 7.Kc2 h4 8.b4 Kg7
  +/-  (0.77)   Depth: 16/52   00:03:37  626mN
1...Qe3 2.Bxb8 Rxb8 3.Bxb5 Qc5 4.Rd5 Qc7 5.Ba4 Qxc3 6.Rb5 Rxb5 7.Qxb5 Qc7 8.Rc1 Qd8 9.Qa6
  +/-  (0.87)   Depth: 17/54   00:05:02  865mN
1...Qe3 2.Bxb8 Rxb8 3.Bxb5 Qc5 4.Rd5 Qc7 5.Ba4 Qxc3 6.Rb5 Rxb5 7.Qxb5 Qc7 8.Rc1 Qd8 9.Qa6
  +/-  (0.87)   Depth: 17/54   00:08:22  1441mN
1...Qe3 2.Bxb8 Rxb8 3.Bxb5 Qc5 4.Rd5 Qc7 5.a4 e4 6.Qc6 Qxc6 7.Bxc6 Rc8 8.Bd7 Rc7 9.Be6 f4
  +/-  (0.82)   Depth: 18/56   00:14:48  2522mN
1...Qe3 2.Bxb8 Rxb8 3.Bxb5 Qc5 4.Rd5 Qc7 5.a4 e4 6.Qc6 Qxc6 7.Bxc6 Rc8 8.Bd7 Rc7 9.Be6 f4
  +/-  (0.82)   Depth: 18/56   00:20:48  2954mN

(,  30.12.2007)
Parent - - By Felix Kling (Gold) Date 2007-12-29 23:36
Jeroen is right. The decisive mistake was 30.... Qh3??- after that Rybka didn't have a chance to survive. Black is already lost in the 37. move, doesn't matter if you play Qe3 or any other move...
Parent - - By Uri Blass (*****) Date 2007-12-30 04:34
You may be right but I am not sure about it.
It only seems that black is lost in move 37 but I do not claim to know what is the losing mistake
I simply want to see when black lost material.

Chess is too complex game to know and I did not analyze it for many hours.

It seems to me that number of pawns is relatively unimportant here because there are many pieces on the board and king attack is more
important.

if you lose rook for a bishop it helps your opponent to get king attack and it also helps your opponent to protect against king attack because you have less pieces to defend.

The only hope of black to stop the white pawn majority is by attacking the white king if the pawns go forward so black should not allow trading the rook b8 that may help him in king attack).

Inspite of it I see that white king attack seem to be too strong also after 37...Rc8.

Uri
Parent - By Venator (Silver) Date 2007-12-30 08:47
The simple point is that a trade of a- and b-pawn for h- and g-pawn is very favourable for white. The a-pawn
is supported by two strong bishops and the white coloured bishop controls a8. So white only needs to go
a3-a4-a5-a6-a7 and the game is over. The black pawns cannot create any counter chances.

So for me the game was decided once Rybka chose to go for this trade....
Parent - - By M ANSARI (*****) Date 2007-12-30 07:35
I agree!  Qh3 looks like the Queen is totally misplaced.  It seems interesting though that Rybka likes its position.  But the Queen is out of play for many moves and the resulting complications favor white.  Still it is not often you see Rybka lose a game where it sees itself with an advantage, so credit should go to the Toga Cluster team for an excellent game.  Complications in very sharp positions usually favor the stronger hardware and credit should be given to the Toga team for steering the game into exactly such a position.  I for one am very impressed ... maybe if Vas is reading this he might start thinking about a similar project :P
Parent - - By Hetman (*****) Date 2007-12-30 08:46
The project would be interesting from IT  but from the market point of view ?
Who from the clients (us) does have computer cluster and testing in such the environment needs investing in hardware ?
The market are 1,2,(4) processors machines. One matter are races (for that you need sponsorship) other matter is being everyday life.
Rgds
Hetman
Parent - - By M ANSARI (*****) Date 2007-12-30 11:20
I personally think that such a project will have a lot of trickle down advantages.  Look at Formula 1 cars ... many of the innovations that first came out there trickled down to every day use ... turbo chargers ... differential slip ... dynamic suspension ... etc.  I believe the same benefit will come from a system where several motherboards are used.  If each motherboard would be tuned to using an engine that specifically does one thing very well, then a combination would cover a lot of area.  As it is, a slight modification in the engine that would work solving one type of position might end up making the engine weaker overall.  Thus it makes sense to improve things without hurting overall performance until a proper mix is found.  For example Rybka Winfinder can solve many tactical positions 10X faster than the best Rybka ... but it is weaker overall.
Parent - By Hetman (*****) Date 2007-12-30 11:48
F1 races are sponsored by companies which have a lot of money or are producing a lot of cars Ferrari-Fiat, Mclaren -Mercedes,
Wiliams-BMW etc. So the rich companies are doing advertising that way or work out the prestige. If some of the chess-programmers get sponsorship of IBM, Sun etc. they may start such a development. If not they have produce that what will be bought by gray people ;-).
You can run Winfinder and Rybka algorithm 'paralelly' on SP even. The OS takes a care about processor assignment in the time slices.

Concerning turobo-charging I am not sure if that has not come from the lorries to the F1 :-).

Rgds
Hetman
Parent - - By Vasik Rajlich (Silver) Date 2007-12-30 12:55
Majd,

as you know, I am thinking about exactly that. This is for after the Rybka 3 release, before the tournaments in the summer.

Vas
Parent - - By M ANSARI (*****) Date 2007-12-30 15:44
Good to hear Vas ... I thought you had forgotten all about it :)  Seriously though ... I do think it will bring chess to new heights.  Tricky tactical and endgame positions will become a thing of the past, and work done on such a project will not be wasted because it will have a perfect fit once cores get to 32 cores +.  The holy grail will be independent brains with expertise in specific areas that share the same hash tables ... although even withought sharing hash, it would still be a huge advance because you can fine tune many different engine parameters to solve one type of position without worrying about it affecting the overall strength of the engine.  And remember that the number of different engines that can be used is limitless.
Up Topic Rybka Support & Discussion / Rybka Discussion / Rybka in Paderborn
1 2 3 4 5 Previous Next  

Powered by mwForum 2.27.4 © 1999-2012 Markus Wichitill