|
< Earlier Kibitzing · PAGE 7 OF 43 ·
Later Kibitzing> |
| May-28-05 | | Montreal1666: <SnoopDogg: So the Hydra vs A Nickel, 2004 game means nothing? Hydra losing to a sub 2600 GM with both pieces black and white means nothing? Oh well.
Anyways if Anand were playing Hydra I don't think Hydra would have a chance. I don't care if those games were correspondence Hydra is all hype judging by those games. Anand would crush it.> This game is against an earlier version of Hydra. It is the latest version that is supposed be very strong (stronger than Deep Blue). I guess we will only after the games against Adams. |
|
| May-28-05 | | WillC21: <SnoopDogg> You realize that GM Nickel had like 10 days per move, right? Snoop, if you and I played against the same version of Hydra, and had 10 days per move for us to play a move with computer assistance and all, we COULD put up a fight. But if we separately played Hydra on classical OTB time controls we'd each get smashed to pieces. See my point? I'm not saying the correspondence games mean "nothing," but they don't mean as much as you are implying. |
|
| May-28-05 | | Dionyseus: Hydra "Chimera" is an old, weak version of Hydra.
The Hydra we saw in the Hydra vs Shredder match and the Hydra+Fritz+Junior vs Topalov+Ponomariov+Karjakin match was 16 clusters. Hydra dominated and remained undefeated.
http://www.chessbase.com/newsprint....
http://chessbase.com/newsdetail.asp...
The Hydra that dominated the 14th Paderborn Computer Chess Championships a couple of months ago was 32 clusters I believe. 7 wins, 2 draws, including a win over Shredder with the black pieces!
http://wwwcs.uni-paderborn.de/~IPCCC/
The Hydra that will be facing Adams this June will be 64 clusters, 200 million positions per second, just as fast if not faster than Deep Blue! Not only is Hydra just as fast, but it has a superior search algorithm and positional knowledge. Adams has no chance at all, and I think Adams is simply one step towards becoming the undisputed champion of chess. |
|
May-28-05
 | | Ron: Kramnik once called a Anand-Topalov game "twenty-third century chess."
Well, if the creators of Hydra are right that their machine plays at 3000 ELO, we might see some twenty-third century chess in the Adams-Hydra match. |
|
| May-29-05 | | THE pawn: As I said in previous posts, Hydra will kick Adams' butt. Gm Nickel played against the weaker version of the hydras, now the one fighting Adams has 64 clusters, and with 32 clusters he was already able to beat Shredder with black. He is undefeated and will beat Gm Adams, period. I know I'm talking like I'm on the side of Hydra, but I like a lot Adams and I hope he'll be able to put up a match...but I have my doubts about this...Still, it's gonna be a hell lot of fun to see this brand-new technological monster with a perfect ELO rating against the great Adams...And for those complaining that it should have been Anand or Topalov, I don't think it would have made any difference, Adams is perfect for the challenge and this will be enough to show the world how strong a machine has become. |
|
| May-29-05 | | SnoopDogg: <Gm Nickel played against the weaker version of the hydras, now the one fighting Adams has 64 clusters, and with 32 clusters he was already able to beat Shredder with black.> GM Nickel is also weaker (alot weaker) than Adams. I don't care how fast it could think. I bet half of the people here know that Fritz 5.32 thought faster than Fritz 8 or Junior 9 does. Yep that's right and no one here will say Fritz 8 is weaker than Fritz 5.32. I'm interested in its chess knowledge not its superior calculating skills. The reason why it beat Shredder (it actually struggled in lots of its draws in the endgame) is because Shredder cannot think like a real GM and although it comes close its just not the same. Topalov had Hydra on the ropes and sticking a developing teenage and some wannabe world champion (Ponomariov) against it and saying its the best ever is just insane. However if it white washes Adams I will give its respect as the best computer out there. |
|
May-29-05
 | | JointheArmy: Well said! |
|
| May-29-05 | | Swapmeet: I would like to think Adams is well prepared to at least draw with white, so I'm hoping its not a total wash. Now is this "Hydra 64" basically the same program that played in that man vs machine tourney, but with better hardware? If so then Adams will at least have some knowledge of how it thinks. If not then my hope rests on Adams, however, my money's on the machine. |
|
| May-29-05 | | andrewGonzales: I dont see the things that way, a draw with white will be a shame, Adams can beat Hydra and im sure he will do, if a 2600 gm can do ti with black, adams will demolish it, even tough the program has improved |
|
| May-29-05 | | Hesam7: How about a FRC match between Hydra and Fischer? |
|
| May-29-05 | | TIMER: Even if it is stronger than Deep Blue, top chess players could beat it. Deep Blue did not play that well more than Kasparov self-destructed. Kasparov would have very likely won a rematch. [Kasparov resigned when he had perpetual to draw, then last game he played opening known to be losing in theory, Deep Blue had to do nothing, Kasparov was psychologically defeated. He should have won 4-2 instead of losing 3.5 2.5] |
|
| May-29-05 | | Chrilly: I am the programmer of Hydra. We play against Adams, because the match is in London and we wanted a "local hero" as our opponent. Besides this I think M.Adams is one of the best - maybe even the best - player against a computer. Adams can play like Adams. Kasparov can e.g. not play like Kasparov. When it comes to tactics, computers are better. One can only beat them "quietly".
Besides this is GM Adam a real gentleman. I think the match will also be a lot of fun for all involved. |
|
| May-29-05 | | WMD: Have you been to the Wembley Conference Centre venue yet? What sort of stage set-up is envisaged? |
|
| May-29-05 | | THE pawn: <SnoopDog>
64 clusters equals to a lot of calculating. The machine's tactics are undeniable, and his virtual knowledge of chess is unmatched. So unless Adams comes out with an incredible strategy involving tons of strange moves, he will lose. |
|
| May-29-05 | | SnoopDogg: <andrewgonzales> <TIMER> I agree with everything you two just said. And Timer that was one of the most intelligent things ever said about the Deep Blue match. To me Deep Blue proved nothing and the whole match was surrounded by shady behaviors on the Deep Blue team. Although Game Over didn't do a great job on the match the movie, it still demonstrated the match was overhyped. |
|
| May-30-05 | | RisingChamp: Are you so sure of that <Chrilly> for example Sergey Karjakin beat deep Junior in the Najdorf recently at Bilbao, of course I agree trying to beat a comp at tactics can be silly,but isnt it possible to emerge from the opening with advantage due to home preparations |
|
| May-30-05 | | Doctor Who: One of my favorite examples of humans beating computers tactically is Kasparov vs Deep Thought, 1989, watch how Garry wins the computer's queen in the opening then cleans house while the computer tries in vein to develop its remaining forces. A lot has changed in 15 years, though, and I doubt that Hydra can be beaten like this. <Chrilly> wonderful to have you on the site. Keep us posted with what's happening with the Hydra project! |
|
| May-30-05 | | Dionyseus: <Chrilly> Welcome to Chessgames.com! I wish your program good luck in both the Adams match and the PAL/CSS Freestyle Tournament. |
|
| May-30-05 | | Chrilly: Well Kasparov against Deep Thought was some time ago. Deep Thought searched about 11 Plies (plus a lot of extensions). Hydra 18 Plies (minus pruning, plus a lot of extensions). In my experience is the normal "search horizon" of GMs 12-14 Plies***. Depends of course on the position. There are some known patterns e.g. for king-attack, were they see quite long. But in chaotic, computer-chess-type positions they overlook relative short tactics. It is also a problem of time and effort. The error rate increases with every move in a sharp position. 10 moves in a row without an error is almost impossible. *** I calculate this from the lag between the announcement of Hydra that it has found a winning combination and the moment when the face of the human turns red (Adrenalin has grasped whats going on). E.g. Hydra finds in move 30) the combination, the face typically turns red in move 32) or 33), sometimes even later. |
|
| May-30-05 | | Chrilly: Opening preperation is certainly a critical point. We use a very short opening book. Maximum 10 moves, sound main variations. Then we leave Hydra from the lash. I think it is difficult to find within the first 10 moves in main-variations killing-novelties. It is possible to find something afterwards. But Hydra is difficult to predict and it has already refuted some lines. Testing the novelty with Fritz is no real guaranntee. If the GM deviates quite early from theory, he has less time per move and has to do much more calculations. I think Adams will therefore be interested to keep the game in known territory.
In case of Karajkin against Junior Junior committed simply suicide. Junior has won numerous games with its typical h-pawn advance, but in this game it was completly wrong. Difficult for a programm, but e.g. Ponomariov made against Hydra also a very severe positional error by exchanging his light squared bishop. Hydra "knew" that this was wrong. The evaluation went up 1/4 pawn. The times they are changing, nowadays it is not always the computer who makes the stupid moves. |
|
| May-30-05 | | GreenMamba: I agree with many people here, Kasparov didnt win the match 4-2 because he played the bad caro-kann opening, that has proved that when white plays correctly agressive style, black pieces get a very difficult game, the classical horse sacrifice or putting up pawns in kings side leads to a very complicated games when black plays this opening. And lets not forget something important, most of the times computers are extremely weak in end games, even tough they can handle openings well most of the times, so sometimes a good Gm can force to an end game and win relatively simple, im sure Adams will win the match, i just hope adams wont repeat kasparov mistake of playing the bad caro-kann. I will like to see those games played by karkajin with junior, i have try to find them but cannot see them, can someone can give me the web site to see them?? i will aprecciate it. |
|
| May-30-05 | | GreenMamba: Does someone knows why Deep Blue was retired??? i guess why IBM didnt kept improving it like Fritz do. |
|
| May-30-05 | | GreenMamba: I just saw the game karkajin win against junior, he outplayed absolutely the computer, it looked like a game between a 2750 Gm versus a beginner, and Topalov also missed a victory against Hydra in this old match, every time im more sure Adams will win the match. |
|
| May-30-05 | | Dionyseus: <GreenMamba> Me and several others are discussing whether or not Topalov had a victory. I believe Topalov did not have a victory. Check out the discussion here: Topalov vs Hydra, 2004 |
|
| May-30-05 | | Chrilly: Topalov-Hydra was a draw. Hydra made a severe positional mistake early in the game, Topalov had a positional won position, but could not win it. He was very disappointed after game "Would have won this against any other human". |
|
 |
|
|
< Earlier Kibitzing · PAGE 7 OF 43 ·
Later Kibitzing> |
Kibitzing
