< Earlier Kibitzing · PAGE 63 OF 79 ·
|Jun-15-09|| ||SugarDom: I'll download chessmaster challenge, the free edition first...|
i dont have it in this computer
|Jun-25-09|| ||Pawnsgambit: Rybka will completely destroy Anands and Kramniks of this world. In a 12 game match Top shot GMs would hardly be able to draw 2-4 games; Rybka has never been beaten by a human in classical time control.|
|Jun-25-09|| ||Pawnsgambit: I would still like to see engines running and beating GMs without table bases and opening books.|
|Jun-25-09|| ||alexmagnus: <Pawnsgambit> I occasionally play in the engine room on Playchess with a Rybka without an opening book and TBs - exclusively against humans. So far it is unbeaten (there were some draws though and no GM played it, but a couple of IMs (all IMs were beaten btw, the draws were all against ~2200s by repetition nearly in the opening).|
|Jul-06-09|| ||Dionyseus: Highendman is the first centaur to beat the Rybka 4 beta cluster:|
Here's the clean pgn:
[Event "Friendly Game, 60m + 30s"]
[Site "Engine Room"]
[White "Highendman, Centaur"]
[Black "Rechenschieber, Rybka 3g55 cluster"]
1. e4 e5 2. Nf3 Nf6 3. d4 Nxe4 4. Bd3 d5 5. Nxe5 Nd7 6. Nxd7 Bxd7 7. O-O Bd6 8.
c4 c6 9. cxd5 cxd5 10. Nc3 Nxc3 11. bxc3 O-O 12. Qh5 f5 13. Qf3 Kh8 14. Bd2 Qa5
15. Rfe1 Rae8 16. a4 Rxe1+ 17. Rxe1 Bc6 18. Bc2 Qb6 19. Bc1 Qc7 20. h4 h6 21.
g4 Qf7 22. g5 hxg5 23. hxg5 g6 24. Bf4 Bxf4 25. Qxf4 Re8 26. Rxe8+ Qxe8 27.
Qe5+ Kh7 28. Bb3 a5 29. Qf6 Qd7 30. Kh2 Qg7 31. Qe5 Qxe5+ 32. dxe5 Kg7 33. Kg3
Kf8 34. f4 b6 35. Kf2 Kf7 36. Ke3 Ke7 37. Kd4 Ke6 38. Bd1 Ke7 39. Bf3 Bxa4 40.
Bxd5 Be8 41. c4 Kd7 42. c5 bxc5+ 43. Kxc5 Ke7 44. Kb6 a4 45. Ka5 Kf8 46. Kb4
Bd7 47. Bf3 Ke8 48. Bd1 Kf8 49. Bxa4 Bc8 50. Kc5 Ke7 51. Bb3 Ba6 52. Bg8 Be2
53. e6 Bf3 54. Bf7 Be4 55. Kd4 Bf3 56. Bxg6 Kxe6 57. Bh7 Bh5 58. Kc5 Bg4 59.
Bg8+ Kd7 60. Kd5 Ke7 61. Ke5 Kf8 62. Be6 Kg7 63. Bxf5 Be2 64. Be4 Kf8 65. f5
Ke8 66. f6 Bc4 67. g6 Bf7 68. g7 Kd7 69. Bd5 Ke8 70. Bxf7+ Kxf7 71. Kf5 Ke8 72.
g8=Q+ Kd7 1-0
According to Highendman, the game is won from move 32.
|Jul-08-09|| ||Isbjorn: Does anyone know if there has been any recent computer chess tournament where the chess programs did not use an opening book? Or what Rybka's estimated rating is when it doesn't use its opening book?|
|Jul-08-09|| ||blacksburg: that would be interesting, computer tournament with no opening books.|
don't worry about rybka, it would still destroy 99.99% of humans with no opening book and giving pawn and move odds. :)
|Jul-11-09|| ||Dionyseus: The Rybka team has issued an open challenge to anyone with a quad or better, or a FIDE rating over 2000, to play against the Rybka 4 beta running on their cluster monster: |
So far the cluster with Rybka 4 beta has lost only one game out of 24 against top centaurs. +9-1=14
|Jul-11-09|| ||SamAtoms1980: I would love to do this, but all I have is a single processor with Fritz 6 running on it. Even if we'd be cannon fodder I'd like to see just what makes Rybka so dominant over every other engine. But if I were to ask, I'd probably be laughed off.|
|Jul-11-09|| ||Dionyseus: The Rybka 4 beta cluster just lost another game, to the same centaur! This is the first time Rybka 4 beta has lost with the white pieces, a great achievement by Highendman:|
[Event "60m + 1s, unrated"]
[White "Rechenschieber, Rybka 3g55 cluster A"]
1. e4 c5 2. Nf3 d6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 a6 6. Be3 e5 7. Nb3 Be6 8. f3
h5 9. Qd2 Nbd7 10. Nd5 Bxd5 11. exd5 g6 12. Be2 Bg7 13. h3 h4 14. O-O O-O 15.
Rfe1 Rc8 16. Rac1 Nh5 17. Na5 Qc7 18. c4 Ng3 19. c5 dxc5 20. b4 Qb6 21. bxc5
Rxc5 22. Nc4 Qd8 23. Bxc5 Nxc5 24. Ne3 Ne6 25. Bd3 Nf4 26. Kh2 Qd6 27. Be4 Rd8
28. Nc4 Qf6 29. Na5 Rd7 30. Rb1 Qg5 31. Nxb7 f5 32. Nc5 fxe4 33. fxe4 Rd8 34.
Nxa6 Bf8 35. Nb4 Rb8 36. a3 Bc5 37. Rb3 Kg7 38. Rf3 Rxb4 39. axb4 Bxb4 40. Qf2
Bxe1 41. Qxe1 Nxe4 42. Qg1 Kh6 43. d6 Nxd6 44. g4 hxg3+ 45. Rxg3 Qd8 46. Rb3
Nf5 47. Qf2 Qd1 48. Ra3 Qd6 49. Qa2 e4 50. Ra6 Qd3 51. Ra3 Qd4 52. Ra8 Nh4 53.
|Jul-28-09|| ||newzild: I drew a blitz game with Rybka once when I turned off the opening book. It played a bad line of the Spanish and I won a pawn. Later I returned the pawn to get a draw by threefold repetition. This was one of only about seven games I've ever played with Rybka, and I'm only rated 2050-odd. So I'm not sure if it would "beat 99.99999 per cent of humans with its opening book turned off."|
|Jul-30-09|| ||aragorn69: Any free version of Rybka available on the net??
(Sorry if this has been asked before: too lazy to scroll back all those pages :-) )
|Jul-30-09|| ||myschkin: . . .
|Aug-02-09|| ||myschkin: . . .
Rybka wins Chess960 Computer World Championship (Mainz)
|Aug-02-09|| ||jhoro: <aragorn69: Any free version of Rybka available on the net??
(Sorry if this has been asked before: too lazy to scroll back all those pages :-) )>|
you can actually download for free Rybka 2.2n2 for free from Rybka Chess directly
|Aug-02-09|| ||kingsindian2006: checking out kibitzing on rybka's page is a nice surprise, its one of the few pages where u dont see people critizing games and saying a computer says try this or that. what a relief on the eyes|
|Aug-03-09|| ||squlpt: <newzild> You'll find that drawing Rybka (while an excellent and rare accomplishment) is much much easier than defeating it|
|Aug-28-09|| ||kooley782: A lot of people say that it is impossible to beat the best computer. It's hard to think of anything much further from the truth. A human at its very best could beat a computer at its very best. No computer is "impossible" to defeat, not even Rybka.|
|Sep-11-09|| ||ex0duz: I was just having an interesting conversation online and someone just said this..|
<Ofcourse there is an optimal play. We just don't know it yet. And no human has any chance at all against Rybka 3 in a tournament game, which has approx. 3300 rating, compared to top human ~ 2800.>
I'm arguing against it, and saying that there is no optimal play yet, because i don't think that the game is solved or can be considered as such with current software/hardware. He's basically saying that we have no chance at all against an approximately 3300 rating machine. Is this true? This conversation is interesting and i thought you guys would know a hell of a lot about it and could share some information with a chess/A.I 'noobie' like myself.
When was the last time a human beat the most powerful programs at classical time?
Can someone give me a list of all the most recent times in competition that humans won? What year, who played what, location, prizemoney, whatever.
Also, whatever other stats you think would be interesting. Like overall percentage of draws, wins, losses, or anything else similar.
Thanks a lot guys. Appreciate it.
|Sep-11-09|| ||zanshin: <ex0duz> You need to take engine ELO ratings with a grain of salt because they are estimated against other engines - the same way human ELOs are calculated by play against other humans. I think it's fair to say that engine ELOs are inflated because the manufacturers use a high baseline for marketing purposes. There are simply not enough engine vs GM matches to get reliable ELO ratings for most engines. Also, all engines tend to have the same shortcomings. They are not addressed because there is no need to (since engines are evaluated against each other).|
Having said that, check out this link:
http://www.rybkachess.com/index.php... It shows some matches between Rybka and human GMs. Note that the GMs are given handicap odds. My opinion is that Rybka would achieve an ELO of about 2800 to 2900 if evaluated against human GMs at tournament time controls (using reasonable hardware and opening book).
|Nov-10-09|| ||whatthefat: <zanshin: My opinion is that Rybka would achieve an ELO of about 2800 to 2900 if evaluated against human GMs at tournament time controls (using reasonable hardware and opening book).>|
It would be a fascinating experiment (one that some might argue Danailov has already done :P). My feeling is that Rybka would do better than that. I would expect it to win at least half of its games and draw the rest, which against 2750 opposition would mean a performance of over 2940. A rating of 3300 on the other hand seems a bit silly. Even a +9 =1 -0 result in a 6 player double round robin would only equate to a 3200 performance.
|Dec-14-09|| ||notyetagm: <HIARCS?> <SHREDDER?> |
Are there any chess engines on the near or even distant horizon that can compete with Rybka?
|Dec-15-09|| ||RandomVisitor: At 40 moves/120 minutes tournament time control, here are the top engines running on quad processors: CEGT list:|
Rybka 3 - Elo 3166
Naum 4 - Elo 3105
Rybka 2.3.2a - Elo 3070
Deep Fritz 11 - Elo 3014
|Dec-15-09|| ||mrandersson: Played a game last night on chessbase played a guy who was a guest but i think he was a cheat. can any 1 else say for sure the other person i played was a guy using a computer? Im not fischer of course but this guy was very comp like lets say.[Event "chessbase"]
e4 b6 Bb7 Bg5 Bxe4 d5 f6 Nc3 Bxg2 Bxg2 fxg5 Qe2 Nf6 0-0-0 a6 Nf3 h6 Qd3 Kf7 Ne5+ Kg8 Rh1-g1 b5 Bh3 g4 Bxg4 Nxg4 Nxg4 b4 Ne4 a5 d6 c6 Qb3+ e6 Ngf6+ Kf7 Qf3 g5 Qh5+ Kg7 Nxg5 Qxf6 Ne4+ Qg5 Nxg5 hxg5 Rxg5 Kf6 Qg6++ 1-0
|Dec-15-09|| ||whatthefat: <mrandersson>
Given Black was completely lost out of the opening it's hard to say. I think it can at least be said that White was a fairly strong player given the accuracy of the finish (17.d6! and 18.Qb3+!) but I don't see any concrete evidence to suggest cheating. What was the time control anyway?
< Earlier Kibitzing · PAGE 63 OF 79 ·