|
< Earlier Kibitzing · PAGE 3 OF 3 ·
Later Kibitzing> |
| Jun-23-05 | | radu stancu: I was just wondering, will Kasim play other games against Accoona, or was it a one-game event? |
|
| Jun-23-05 | | SCapp123: I was present at this exciting one game match in a midtown NY TV studio. It was played with a time limit of one hour + 10 second bonus per move for each side. At the end, Kasimdzhanov had used up almost all of his time reserve, the computer considerably less. Kasim. attributed his willingness to take a draw to his time situation. Some GM's present were confident that without the time problem, white would go on to win, but some others there thought the machine might be able to put up a defence. In any case a representative of the programming team remarked that it was impressively bold of Kasim. to make a knight sacrifice against this powerful program. |
|
Jun-23-05
 | | WannaBe: <radu> I believe it was just one game. <Scapp123> Duplicate post, do you have time to delete one of them? |
|
| Jun-23-05 | | nikolaas: Does anybody know or the programmers include a penalty for closed positions? Considering that computers are worse at them such a system seems logical to me. |
|
| Jun-23-05 | | aw1988: The sacrifice was certainly brave, (and brilliant!) and the game quite interesting. |
|
| Jun-23-05 | | ConspTheory06: <al wazir> I was thinking the same thing how does black respond to 39.Rf4? |
|
| Jun-23-05 | | sitzkrieg: I think black is forced to play ..g6 after wich Rh4 is useless (Nh7). So white takes the pawn at g6 and has three pawns and good position for the piece = equal.. |
|
| Jun-23-05 | | sitzkrieg: uhmm my mistake..not true; pawn at a5 falls too. Still bisshops opposite colors, and nice pawns at h,e and f lines good enough for white i guess |
|
| Jun-23-05 | | aw1988: How about 34. Bg6? |
|
| Jun-24-05 | | patzer2: Rustam's 32. Ne6!! is indeed a deep positional sacrifice, demolishing Black's Kingside pawn structure with the sacrifice on e6, as part of a neat defensive combination, and guaranteeing White at least an even game. Note that in the final position, Black must accept a draw by threefold reptition of moves with 42...Rg8 or concede White a clear to winning advantage with other choices. |
|
| Jun-24-05 | | patzer2: Rustam's followup defensive move 40 Rg5! forces 40...Rg8 with a draw by threefold reptition of moves quickly following. If Black tries to vary from the forced 40...Rg8 reply, he loses. For example, after 40... Bxa5 41. Rf4 Bd8 42. Rh4+ Kg8 43. Rh6 Rff7 44. Kh1 Rbc7 45. Qg6 Kf8 (45... Rfe7 46. Rgh5  ) 46. Rh8+ Ke7 47. Rxe8+ Kxe8 48. exf7+ Rxf7 49. Bf5 Kf8 50. Be6 Rb7 51. Rf5 Ra7 52. Qh5 Ba5 53. Qh7 Ke7 54. Rxf6 Kxf6 55. Qf5+ Ke7 56. Qf7+ White has a decisive advantage. |
|
| Jun-24-05 | | kevin86: I guess this is a gotd because of the present Adams-Hydra,human-computer match. Yes,and the pun was irresistable--the game was a real snoozer :( |
|
| Jun-25-05 | | vonKrolock: Monster? Accoona play like a PSOne CD... |
|
| Jun-25-05 | | csmath: Notice that Alcoona is "off the shelf laptop." Frankly I don't know what that means precisely as it is not specified anywhere that I know. One thing you can be sure is that Hydra calculate many multiple more nodes and goes into significantly deeper analysis than a laptop computer. Adams has much harder job than Kasim had. My best bet is that Kasim would not be any better against Hydra than Adams. |
|
Jun-25-05
 | | Honza Cervenka: <One thing you can be sure is that Hydra calculate many multiple more nodes and goes into significantly deeper analysis than a laptop computer.> I don't doubt much about it but this doesn't change anything on my suggestion that a human player can be successful against any of today's top computers if he manages to create such a positions in which any "deep analysis" simply doesn't matter. On the other hand, in positions where the deep comp analysis matters the first "human" mistake is almost certainly decisive one and it is not so important if the computer playing against the man is Hydra the Great or "only" Accoona Toolbar.:-) |
|
| Jun-25-05 | | Anna Wrekzia: Is the premise behind this "toolbar" thing that you can simply download a program this strong and have it play on a PC? Or is this just a product promotion thing, and has nothing to do with the computer program...? |
|
| Jun-25-05 | | csmath: <<I don't doubt much about it but this doesn't change anything on my suggestion that a human player can be successful against any of today's top computers if he manages to create such a positions in which any "deep analysis" simply doesn't matter.>> Well, that is true but that is a problem as well. It is hard to generate better position against a more powerful machine. Even on a single processor few hundreds of MHzs make a lot of difference. Here we are talking about supercomputer, dedicated to play chess. That is a huge difference. Laptop that Kasim played against is probably <2500 strong, Hydra is 3000 strong. The difference is huge. 2500 player cannot hope to win a game against Kasparov much less against Hydra. (I think Hydra is stronger than Kasparov). Take a look at the game #3. That is a positional masterpiece with flawless execution. Adams had no chance. Only in the game #2 he made something of a decent position but was afraid to force any attack and eventually made an error and barely survived after succeeding in a blockade. Perhaps we need to see Leko playing here but again keep in mind this is the most powerful hardware any of chess elite GMs has ever faced. Kasparov's rival Fritz in 2003 was much, much weaker. |
|
| Jun-28-05 | | s4life: Hydra played the best positional chess I've seen yet, human or computer, with an ELO performance of over 3000 - after all positional play is just fancy jargon for a thorough understanding of all the variations, which in case of computers is directly proportional to the amount of calculations per second.
BTW, Hydra is far from being a supercomputer (it's just that our national labs are not concerned in defeating chess GM's). it's only a 32 processor cluster... even without hardware acceleration, I am pretty sure the 1000 processor cluster running in my school would be way faster; now imagine what a 20 million dollar craig supercomputer would do to any GM |
|
| Jul-01-05 | | Wilhelm: Yet-another J Nunn's modified (Accoona version) perspective for the Future of Chess: June 2005: Kasim draws with the Toolbar.
Dec 2007: Nakamura draws with the 'Edit' option.
Feb 2009: Karjakin draws with the 'Minimize' button.
Sept 2011: Carlsen draws with the Mouse pointer. |
|
| Jul-01-05 | | PhilFeeley: No one answered <al wazir>'s question about 39. Rf4, so I played with it for a while. 39. Rf4 g6
40. Rh4+ Rh7
and then what? Is black safe?
41. Rxh7 Kxh7
then what? 42. Qg3?
If:
40. Rxg6 Rxg6
41. Qxg6 Qxg6
42. Bxg6 ...
and does white have enough pawns? It looks like he does. (If 39. Rf4 Rg8 White should have a forced win with: 40. Rh4+ Kg8 (nothing else)
41. Rgh5 g6
(41....Nxh5 42. Qh7++ 41....Qxh5 42. Rxh5 Nxh5 43. Qh7++) 42. Rh8+ Kg7
43. Rxf8 and if
43....Kxf8
44. Rh8+ winning the queen and mate in a couple more moves. or 43...Qxf8
44. Qxg6++)
I'm sure I've missed something. I usually do! |
|
| Aug-01-05 | | crafty: 39. ♖f4 g6 40. ♕f3 ♘h7 41. ♖g1 ♖bg7 42. ♖fg4 g5  (eval -0.66; depth 16 ply; 1000M nodes) |
|
| Aug-11-05 | | BaranDuin: How come the toolbar drew the world champion while I beat it five times in a row on the hardest level. |
|
| Aug-11-05 | | Koster: <I don't doubt much about it but this doesn't change anything on my suggestion that a human player can be successful against any of today's top computers if he manages to create such a positions in which any "deep analysis" simply doesn't matter.> That is probably the only chance except maybe for Anand whose speed and accuracy of calculations I believe compares with any of the computers. He would be the only current GM i would consider betting on vs. hydra. |
|
| Aug-15-05 | | shortsight: <BaranDuin> Where did you get to play the Accoona Toolbar? |
|
| Aug-16-05 | | jamesmaskell: You can download the toolbar off the Internet and play it there. I played it once. Couldnt be bothered after that game. Deleted it. |
|
 |
|
|
< Earlier Kibitzing · PAGE 3 OF 3 ·
Later Kibitzing> |
Kibitzing
