Kramnik v/s Fritz

Originally posted by zebano
He also has days to consider each move...

That said, he may know something about it's playing style that Kramnik doesn't. However, you should remember that this is DEEP fritzy! (oh lala) NOT THE FRITZ 9 Weyerstrass is used to.

I say we have a Weyerstrass vs. Deep Fritz correspondence match and see if he can fare better than Kramnik. A correspondence GM was able to beat Hydra, which I've heard is more powerful than Deep Fritz, so as a correspondence IM Weyerstrass should have a chance.

Originally posted by buddy2
One unexpected result of the match: The result was actually carried on American TV news. So, all publicity must be good. Actually, i think this is the tipping point of computer chess. Computers are advancing so rapidly. In five years nobody in the world will have a chance against them. As i stated before, Kramnik, playing in his most bland, vapid style ...[text shortened]... crupulous, carbon-based humans from using these machines as an aid in tournaments and matches.

Actually, Kramnik playing in his "most bland, vapid style" completely outplayed Fritz in Game 2 and missed an easy win in the endgame, as Seirawan's analysis has shown. Then he had a simple draw in Game 3, but blundered. The main reason he lost Game 6 was because he took huge risks to imbalance the position and play for a win.

If Kramnik had been in better form he wouldn't have lost those 2 games and his technique would normally have converted his one winning position. So if he had made better use of his chances he could have won the match! This is despite the fact that he has a medical condition that affects his play and is still recovering from his very stressful and exhausting match against Topalov.

However, although Kramnik is the undisputed World Champion, I doubt that he is the right person to be representing mankind. We need to put forward the player who has the best results against machines.

This is quite likely to be an obscure GM or IM who has a lot of experience playing against computers and has worked out the best way to beat them. The high profile GM's who are normally chosen to play these matches are too used to playing against humans and cannot adjust their style sufficiently. Unfortunately what normally happens is that the engine manufacturers are the sponsors and they choose the players.

It would be so much better if the sponsor and match organisers were independent. Why can't one of the millionaires or rich corporations who routinely sponsors tournaments, put some money into a human versus computer match instead? Better still, they should organise a series of preliminary matches in order to find out which human player is best at playing machines.

Finally when we know who is the best player to represent humans, there should be a match for the Overall World Championship!

a very good point,Speelman said that Kasparov was the wrong player to take on Deep Blue for exactly those reasons and Speelman said K not only uses computers himself but has possibly too much respect for them and tried to beat it by out calculating it which is a bit of a non starter.

Originally posted by David Tebb
Actually, Kramnik playing in his "most bland, vapid style" completely outplayed Fritz in Game 2 and missed an easy win in the endgame, as Seirawan's analysis has shown. Then he had a simple draw in Game 3, but blundered. The main reason he lost Game 6 was because he took huge risks to imbalance the position and play for a win.

If Kramnik had been in b ...[text shortened]... st player to represent humans, there should be a match for the Overall World Championship!

Presumably the manufactures of chess engines consult various GMs in developing their programmes. It would be fair to say that those GMs have a far better knowledge of how the engines evaluate positions and would therefore be better equiped to beat them.

Originally posted by Dragon Fire
Presumably the manufactures of chess engines consult various GMs in developing their programmes. It would be fair to say that those GMs have a far better knowledge of how the engines evaluate positions and would therefore be better equiped to beat them.

Yeah, maybe someone like GM Joel Benjamin, who helped the Deep Blue team to beat Kasparov. Although Benjamin probably spent far more time looking through Kasparov's games in order to discover his weaknesses and develop a strategy to defeat Kasparov than he did with Deep Blue.

The problem is finding someone who is up to date with his knowledge of the current top engines but doesn't have financial links with the engine manufacturers. Although I suppose a GM who has worked very closely with one engine could be trusted to do his best to defeat a rival engine. But most of the top commercial engines are owned by Chessbase, so the competition between them is artificial.

I would prefer a truly independent GM challenger, who has no connections to any of the engine manufacturers. In the past there was the IM David Levy, who won a series of matches in the 70's and 80's against engines for various bets against their programmers. Levy was not s particulary strong player but he was a computer expert who had spent a long time developing a style of play to beat engines. Of course todays engines are vastly more advanced than when Levy played them. So it would have to be someone with the same sort of knowledge but a better player.

Did you mean Seirawan's analysis in Game 1, David? I took a cursory look at it (game 1 Seirawan's analysis in Fritz 10), to see if, as Seirawan wrote the win "is as clear as a sunny day." My Fritz did not make S's predicted moves. It played 31. ...b5 and went for K's kingside pawns. In each of the variations I saw, the machine either queened one move ahead of K (probably a draw, tho not positive machine could pull it out) or broke through the kside before kramnik could queen. Have you looked at all the variations on your machine. I'm not convinced Seirawan is right. I'm sure the machine, crunching twenty moves down the line would have picked that up. In one case, after white scoops up the bpawn, Fritz made a neat little sac of its bishop to get a kside pawn through. Really quite instructive end game. I'm surprised nobody's tested Seirawan's analysis. maybe there's something i miseed. Anybody take a look at it?

Originally posted by buddy2
Did you mean Seirawan's analysis in Game 1, David? I took a cursory look at it (game 1 Seirawan's analysis in Fritz 10), to see if, as Seirawan wrote the win "is as clear as a sunny day." My Fritz did not make S's predicted moves. It played 31. ...b5 and went for K's kingside pawns. In each of the variations I saw, the machine either queened one move ahea ...[text shortened]... Seirawan's analysis. maybe there's something i miseed. Anybody take a look at it?

they tested it with the fritz crew back then, and after 15minutes (if I remember right) fritz suddenly realized yasser was right.

If I rember rightly Yasser was on the Deep Blue team that beat Kasparov...maybe he should represent humankind

If Yasser tried, It'd be interesting to see how he scored. I think he should challenge and be vociferous about beating the damn thing. Money would be raised (think of all the publicity) and Yasser could get rich! That's what we need: a grudge match against the machine.

If Wormwood is correct, the programmers better start to get to work. Onward and upward for the machines!