|
< Earlier Kibitzing · PAGE 1 OF 2 ·
Later Kibitzing> |
May-27-05
 | | WannaBe: Black's Rook on a8 can't leave 8th, otherwise white R goes to 8th and mate. g5!! forces black to lose the knight. |
|
| May-28-05 | | roni.chessman: Is this the version of the hydra that plans on playing Adams? Or the Hydra Chimera one? If it is being toppled by a near 2600 player, Adams has a very good chance of smashing this silicon opponent. But of course can't be to rash since Topalov tried to barbecue it but ended up with a draw instead. |
|
| May-28-05 | | roni.chessman: With some research I realized how silly I am...This is (duh) the chimera, a much weaker version and the GM had 10 days to think per move. My bad... |
|
| May-31-05 | | halcyonteam: much weaker hydra?? how much weaker? |
|
May-31-05
| | WannaBe: 32 (old) versus 64 (new) there are currently 9 pages worth of discussion going on at Hydra (Computer) |
|
| Aug-01-05 | | Vaguerant: The GM was also assisted by Shredder and/or Fritz and friends. |
|
| Feb-17-06 | | RandomVisitor: Black's problems begin around move 28 after the Queens come off the board. Computer analysis using Rybka shows that the position slightly favors White at this point, and remains this way for the rest of the game. Can Black improve? One idea is 24...Qb6 25. Bxc5 Rxc5. Or possibly 26...Rfd8 fxg7. |
|
| Feb-17-06 | | ughaibu: Being a computer, wouldn't black choose the same move 24 next time too? |
|
| Feb-17-06 | | RandomVisitor: <ughaibu>Well, Rybka and Hydra are both computer-based systems, but they evaluate a position differently. In fact, I have been spending the last couple of months looking at how Hydra evaluates a position. Rybka and Hydra have different rules for calculating the next move and how much better or worse white or black stands. Hydra has not performed as well in correspondence games as it has in regular games, and it seems that a tiny lack of precision in positional understanding is one of the reasons. Sometimes computers agree, and sometimes they disagree. |
|
| Feb-17-06 | | last pawn: hello |
|
| Feb-18-06 | | RandomVisitor: <last pawn>Welcome to chessgames. |
|
| Feb-22-06 | | RandomVisitor: Here is an analysis of this game:
1. e4 c5
2. Nf3 e6
3. d4 cxd4
4. Nxd4 Nc6
5. Nc3 Qc7
6. Be3 a6
7. Qd2 Nf6
8. O-O-O Bb4
9. f3 Ne5
10. Nb3 b5
11. Qe1 Be7
12. f4 Nc4
13. e5 Ng4 0.27/19
14. Bd4 0.27/19 [last book move]
14...O-O $14 0.36/19
(♖ybka 1.01 ♗eta 13b: 14... Bb7 15. a4 $14 0.27/19) 15. Bd3 $11 0.00/19
(♖ybka 1.01 ♗eta 13b: 15. Bxc4 bxc4 16. Nd2 $14 0.36/19) 15... Bb7 0.01/19
16. Qe2 0.01/19 Nh6 -0.02/19
17. Nd2 -0.01/19 d5 -0.01/19
18. g4 -0.12/19 Rac8 -0.13/19
19. Rhg1 0.12/19 b4 0.15/19
20. Ncb1 $11 0.00/19
(♖ybka 1.01 ♗eta 13b: 20. Na4 Qa5 $11 0.15/19)
20... a5 0.04/19
21. Nf3 $11 -0.24/19
(♖ybka 1.01 ♗eta 13b: 21. b3 Nb6 $11 0.04/19)
21... a4 -0.13/19
22. f5 -0.05/19 Bc5 -0.04/19
23. Qf2 $11 -0.24/19
(♖ybka 1.01 ♗eta 13b: 23. b3 Bxd4 $11 -0.04/19)
23... a3 $15 -0.35/19
(♖ybka 1.01 ♗eta 13b: 23... Bxd4 24. Qxd4 a3 25. bxa3 $11 -0.24/19) 24. bxa3 -0.20/19 Bxd4 $11 0.00/19
(♖ybka 1.01 ♗eta 13b: 24...Qb6 25. Bxc5 Rxc5 $11 -0.20/19) 25. Qxd4 $11 -0.10/19
(♖ybka 1.01 ♗eta 13b: 25. Nxd4 bxa3 $11 0.00/19) 25... bxa3 -0.01/19
26. f6 0.00/19 Ba6 $11 0.02/19
(♖ybka 1.01 ♗eta 13b: 26... Rfd8 27. fxg7 $11 0.00/19) 27. Bxc4 0.04/19 Qxc4 0.25/19
28. Qxc4 0.25/19 Bxc4 0.25/19
29. Nd4 0.25/19 Ra8 0.25/19
30. Rg3 0.25/19 Bxa2 0.25/19
31. Nxa3 0.25/19 Rfb8 0.44/19
32. h3 0.44/19 Rb7 0.50/19
33. Rdd3 0.52/19 gxf6 $16 0.80/19
(♖ybka 1.01 ♗eta 13b: 33... Kh8 34. Rc3 gxf6 $14 0.52/19) 34. exf6 0.80/19 Rb6 0.95/19
35. Rc3 0.95/19 Kh8 $18 1.68/19
(♖ybka 1.01 ♗eta 13b: 35... e5 36. Nc6 Re8 $16 0.95/19) 36. Rc6 1.68/19 Rb4 1.92/19
37. Nab5 1.92/19 e5 2.15/19
38. Nd6 2.15/19 exd4 2.15/19
39. Ra3 2.15/19 Rb1+ $18 3.15/2
(♖ybka 1.01 ♗eta 13b: 39... Rbb8 40. Rxa8 Rxa8 41. g5 d3 42. c3 Kg8 43. gxh6 Rf8 44. Rb6 Bc4 45. Nxc4 dxc4 $18 2.15/19) 40. Kd2 3.15/19 Rab8 3.78/19
41. Rxa2 3.78/19 Rf1 3.78/19
42. Rb6 3.78/19 Rg8 $2 $18 7.09/19
(♖ybka 1.01 ♗eta 13b: 42... Rf8 43. Ke2 Rxf6 44. g5 Rxd6 45. Rxd6 Nf5 46. Rxd5 Re8+ 47. Kf2 Ng7 48. Ra7 Ne6 49. h4 $18 3.78/19) 43. g5 7.09/19 1-0
|
|
| Feb-22-06 | | ughaibu: Presumably Hydra is stronger than Rybka, so in a game that Hydra lost to a human wouldn't the student want the analysis of the human who demonstrated himself stronger than the computer, rather than analysis by a weaker computer? How can one trust Rybka as a critic of Hydra or is the intent to show the various eccenticities of these machines? |
|
| Feb-22-06 | | RandomVisitor: <ughaibu>Ideally, we should just repost Nickel's analysis of this game, and let that stand. I am posting Rybka's analysis here as a curiosity. The numerical score after each move gives a feeling for how the game is going. The places in the game where Rybka and the players differ is a point for discussion. Also, I am testing Rybka's abilities as a general purpose analysis tool. |
|
| Feb-22-06 | | ughaibu: Okay, that all sounds fair enough. |
|
| Feb-22-06 | | RandomVisitor: This link:
http://amici.iccf.com/issues/issue_...Lets you see Nickel's analysis of this game.
24...Qb6 is a possible improvement for Black, and Nickel provides this analysis: 24...Qb6
25.axb4 Qxb4 26.f6 Qb2+ (26...Bxd4?! 27.Qxd4 Ra8 28.Qc3 Qc5 29.Nd4 
and Black's counter play at the queen-side is too slow in face of the strong threat g4-g5.)
27.Bxb2 Bxf2 28.Bd4! and Black would not enjoy winning the exchange. However, Black has 28...Bxd4 Nxd4 29. Nxe5 with possibly a slight advantage. |
|
| Sep-02-06 | | MagnaPsygnosis: Hydra is the strongest in a machine vs. Machine match, while Rybka is the Strongest Program in Human vs. machine match |
|
| Sep-04-06 | | RandomVisitor: Once again, see Nickel's analysis above, but here is a 21-ply analysis of the game: 1. e4 c5 2. Nf3 e6 3. d4 cxd4 4. Nxd4 Nc6 5. Nc3 Qc7 6. Be3 a6 7. Qd2 Nf6 8. O-O-O Bb4 9. f3 Ne5 10. Nb3 b5 11. Qe1 Be7 12. f4 Nc4 13. e5 Ng4 14. Bd4 last book move 14...O-O 0.13/21
15. Bd3 0.00/21 (15.Bxc4 bxc4 16.Nd2 d5 17.exd6 Qxd6 18.Qe4 f5 19.Qxa8 Qxd4 20.g3 Ne3 21.Qc6 Qc5 0.13/21) 15...Bb7 0.00/21
16. Qe2 0.00/21 Nh6 0.00/21
17. Nd2 0.00/21 d5 0.00/21
18. g4 0.00/21 Rac8 0.00/21
19. Rhg1 0.00/21 b4 0.00/21
20. Ncb1 0.00/21 a5 0.00/21
21. Nf3 0.00/21 a4 0.00/21
22. f5 0.00/21 Bc5 0.00/21
23. Qf2 -0.26/21 (23.b3 Bxd4 24.Nxd4 Nxe5 25.fxe6 Ng6 26.Qe3 fxe6 27.Nxe6 Rfe8 28.Bxg6 hxg6 29.Nxc7 Rxe3 0.00/21) 23...a3 0.00/21 (23...Qb6 24.f6 a3 25.b3 Bxd4 26.Nxd4 Nxe5 27.fxg7 Rfd8 28.Be2 Nc6 29.Rgf1 Rd7 30.Nxc6 -0.26/21) 24. bxa3 0.00/21 Bxd4 0.00/21
25. Qxd4 0.00/21 bxa3 0.00/21
26. f6 0.00/21 Ba6 0.10/21 (26...Rfd8 27.fxg7 Ba6 28.Bxc4 dxc4 29.Qc3 Rxd1+ 30.Kxd1 Qb6 31.Ke2 Qb2 32.Kf2 Bb7 33.h3 0.00/21) 27. Bxc4 0.10/21 Qxc4 0.12/21
28. Qxc4 0.12/21 Bxc4 0.12/21
29. Nd4 0.12/21 Ra8 0.17/21 (29...Bxa2 30.Nxa3 Rc5 31.h3 Rfc8 32.Rg3 Bc4 33.Nxc4 Rxc4 34.Re3 gxf6 35.exf6 Kf8 36.Kb2 0.12/21) 30. Rg3 0.14/21 Bxa2 0.12/21
31. Nxa3 0.13/21 Rfb8 0.22/21 (31...Ra6 32.fxg7 Kxg7 33.Kd2 Ra4 34.Nab5 Bc4 35.Nd6 Ng8 36.g5 Ne7 37.Rf3 Ba6 38.Rf4 0.13/21) 32. h3 0.22/21 Rb7 0.22/21
33. Rdd3 0.22/21 gxf6 0.34/21 (33...Kh8 34.Rc3 Rab8 35.fxg7+ Kxg7 36.Rc5 Ra7 37.Nab5 Ra5 38.Ra3 Rxa3 39.Nxa3 Rb4 40.c3 0.22/21) 34. exf6 0.35/21 Rb6 0.77/21 (34...Kh8 35.g5 Nf5 36.Nxf5 exf5 37.Rdf3 h5 38.Rxf5 Kh7 39.Rf4 Kg6 40.h4 Re8 41.Kd2 0.35/21) 35. Rc3 0.77/21 Kh8 0.87/21
36. Rc6 0.88/21 Rb4 1.03/21
37. Nab5 1.03/21 e5 1.35/21 (37...Rc4 38.Nd6 Rxd4 39.Ra3 Rb8 40.Rxa2 Rf4 41.Rb6 Rf8 42.g5 Nf5 43.Nxf7+ Kg8 44.Ne5 1.03/21) 38. Nd6 1.35/21 exd4 1.36/21
39. Ra3 1.36/21 Rb1+ 1.76/5
40. Kd2 1.76/21 Rab8 2.22/21
41. Rxa2 2.31/21 Rf1 2.31/21
42. Rb6 2.29/21 Rg8 5.94/21 (42...Rf8 43.Ke2 Rxf6 44.g5 Re6+ 45.Kd3 Rxd6 46.Rxd6 Nf5 47.Rxd5 Ng7 48.Rxd4 Ne6 49.Rg4 2.29/21) 43. g5 5.94/21 1-0
|
|
| Sep-30-06 | | sharpnova: <MagnaPsygnosis: Hydra is the strongest in a machine vs. Machine match, while Rybka is the Strongest Program in Human vs. machine match> no hydra is stronger in Human vs. machine chess than rybka. this is obvious but i have spoken with the creat of hydra. he expressed his disapointment that hydra wasn't destroying rybka like it was destroying fruit and fritz and shredder.. but that the performance superiority was marked. i think he mentioned 100 elo. you seem to be assuming that since hydra lost against a human and rybka generally doesn't, rybka is stronger against people than hydra. however, this game was correspondence. no program for a long time is going to be able to defeat a correspondence GM. (i doubt even draw for some years) |
|
| Nov-03-06 | | MagnaPsygnosis: <sharpnova>
mmm, spectical but interresting... |
|
| Jul-28-07 | | unixfanatic: Rybka's play is much, much more human-like than any other computer I've seen - it values all sorts of things differently, like the importance of fianchettos, pawn levers, etc. |
|
| Jul-17-08 | | jeeky1996: Nice victory.
In my opinion, all computers are strategically dumb except Rybka of course. |
|
| Jul-24-08 | | ravel5184: <last pawn> Are you the person I think you are? |
|
| Aug-25-08 | | andrewjsacks: This version or level of Hydra is obviously not that strong. |
|
| Aug-25-08 | | dabearsrock1010: towards the 35th moves of the game while white's king appears more exposed black's king is the real glaring weakness on the board which eventually decides the game. The refutation of 37...e5 is stunning and is a result of the king's weakness. |
|
 |
|
< Earlier Kibitzing · PAGE 1 OF 2 ·
Later Kibitzing> |
Kibitzing
