Yeah, it seems unlikely that he can actually pull a win considering that tactical solutions are unlikely to crop up. Fritz is too good tactically. This only leaves subtle positional wins and this is really difficult. Still, Kramnik has shown that the top human can indeed keep pace with a great chess engine. Sure, Kramnik missed the mate in 1, but it should have been easily drawn. The score may not be 2:2, but that's how I would really think of it.
On all the games so far, I haven't seen Kramnik have an inkling of an attack. He just plays defence, exchanging pieces, getting the position down to something bland, sterile with no hope of anybody winning. Fritz is not Topalov. It won't unbalance the position. It doesn't play hope chess. It simply looks at all variations 20 moves down the line and picks the one that has the highest evaluation.
Originally posted by buddy2Yeah, makes one feel like he's afraid of the darn thing.
On all the games so far, I haven't seen Kramnik have an inkling of an attack. He just plays defence, exchanging pieces, getting the position down to something bland, sterile with no hope of anybody winning. Fritz is not Topalov. It won't unbalance the position. It doesn't play hope chess. It simply looks at all variations 20 moves down the line and picks the one that has the highest evaluation.
Originally posted by exigentskyIf Seirawan's analysis is correct, and it looks sound to me, then Kramnik could have won game 1. So the score could be 2½-1½ in Kramnik's favour and it appears the machine is beatable when playing black.
Yeah, it seems unlikely that he can actually pull a win considering that tactical solutions are unlikely to crop up. Fritz is too good tactically. This only leaves subtle positional wins and this is really difficult. Still, Kramnik has shown that the top human can indeed keep pace with a great chess engine. Sure, Kramnik missed the mate in 1, but it should have been easily drawn. The score may not be 2:2, but that's how I would really think of it.
What's Kramnik up against? I got the new Fritz 10 (not deep--single processor). I put Fischer "Game of the Century" in it when, on move 11. he played Na4 against Byrne. It's given two exclams in most books. Fritz found it almost instantaneously. We have reached the point in computers when the very best humans can only hope for a draw--if they play perfectly. Even Crafty was looking 16 ply deep when evaluating position with Kramnik v. Deep Fritz. I would guess Fritz is looking around 20. I'm afraid finding a plan (since computers don't really "plan"😉 is not enough to compensate in a real game with the number crunching Fritz uses. And it can only get worse because the computers will increase ability and human's can't. Unless we wait another million years for evolution to improve our nervous system.
Originally posted by buddy2I'm convinced of the fact that the chess playing programs will influence, change and eventually improve the way humans play. Especially the ways the programs attack will be deeply analysed and implemented by humans.
What's Kramnik up against? I got the new Fritz 10 (not deep--single processor). I put Fischer "Game of the Century" in it when, on move 11. he played Na4 against Byrne. It's given two exclams in most books. Fritz found it almost instantaneously. We have reached the point in computers when the very best humans can only hope for a draw--if they play perfect 't. Unless we wait another million years for evolution to improve our nervous system.
At this stage the humans are finding ways to not be beaten by the strong programs, meaning humans are finding and improving ways to defend against the machine. But after they figured out how to do this they will study the machine's attacking qualities and then ...... ahaaaaa ... they will beat the monster again ..... because they learned from it how to go about.
After the programmers improved their programs the cyclus starts anew ....
I don't think computers "attack" in the normal sense. They look at ALL the possiblities, whittle them down to the most likely, look down theline 15-20 moves, say, and evaluate the position. If the position is a checkmate, it'll go for that. If the best it can do is win a pawn or get a better position it'll go for that. anyway, we'll talk in ten years when they have ten piece table bases, the evaluation has improved, and the speed of computers has improved. I mean Kramnik has the computer's "book." How much of an advantage can he get? Maybe if theygive him a bat and he's allowed to pummel the machine when hefeels the need. I remember about twenty years ago David Levy offered a prize for any machine that could beat a grandmaster. If that offer was still good, he'd be a poor man. Never bet against the machine.
I think you have to play the computers game and not your own. Force it's pieces around with multiple games within the one game. It's hungry and you should be able to predict where it will move it's pieces to react with your moves. Cheat it into an area to free your own pieces and attack in another area....
Predictability is it's weakness.