Who could beat the top programs?

Discussion of anything and everything relating to chess playing software and machines.

Moderator: Ras

glorfindel

Who could beat the top programs?

Post by glorfindel »

According to this article from Chessbase
...One interesting new player is Chinese GM Wang Yue, who has not lost a single game for nine months now and is unbeaten in 82 games (after round nine of the Chess Olympiad in Dresden French GM Etienne Bacrot emerged from a game against Wang shaking his head: "That guy really cannot be beaten!"). Wang, who is number 11 in the world today, is close to breaking Vladimir Kramnik's record of 82 games in 1999 to 2000, and that of Mikhail Tal, who played 93 games without defeat in 1973/74...
I think that if one human still has chances to tie a match with the top programs on top hardware, it must be someone like Wang Yue.

The 82-game undefeated streak means that he has succeeded in eliminating blunders, and this is probably the most important thing.
Furthermore, he must be young, not only in order to have good calculating ability, but also because he must be a child of the computer era. That is, he must have learnt chess through analysing with the computer from the very beginning so as to know his opponent as well as possible. On the other hand, someone may argue that even older players have learned many things from computers and have adapted to their style, so this might not be so important.In any case, the human must be in top form during the event and he must play slightly conservatively.

Carlsen, for example, would have to change his style a little and take less risks, because he has recently tried to push to hard for a win sometimes, and he has lost as a result. That said, Carlsen would be a good candidate, of course, together with Radjabov and Wang Yue. Because it goes without saying that the human should be one of the very best.

It seems like there is diminishing interest in these matches, probably because they are considered lost for the humans. But I think in the last two big matches, the players participating (Kramnik and Adams) were wrong choices as far as picking the most dangerous opponent is concerned. So somebody else could still have chances.
Dann Corbit
Posts: 12792
Joined: Wed Mar 08, 2006 8:57 pm
Location: Redmond, WA USA

Re: Who could beat the top programs?

Post by Dann Corbit »

glorfindel wrote:According to this article from Chessbase
...One interesting new player is Chinese GM Wang Yue, who has not lost a single game for nine months now and is unbeaten in 82 games (after round nine of the Chess Olympiad in Dresden French GM Etienne Bacrot emerged from a game against Wang shaking his head: "That guy really cannot be beaten!"). Wang, who is number 11 in the world today, is close to breaking Vladimir Kramnik's record of 82 games in 1999 to 2000, and that of Mikhail Tal, who played 93 games without defeat in 1973/74...
I think that if one human still has chances to tie a match with the top programs on top hardware, it must be someone like Wang Yue.

The 82-game undefeated streak means that he has succeeded in eliminating blunders, and this is probably the most important thing.
Furthermore, he must be young, not only in order to have good calculating ability, but also because he must be a child of the computer era. That is, he must have learnt chess through analysing with the computer from the very beginning so as to know his opponent as well as possible. On the other hand, someone may argue that even older players have learned many things from computers and have adapted to their style, so this might not be so important.In any case, the human must be in top form during the event and he must play slightly conservatively.

Carlsen, for example, would have to change his style a little and take less risks, because he has recently tried to push to hard for a win sometimes, and he has lost as a result. That said, Carlsen would be a good candidate, of course, together with Radjabov and Wang Yue. Because it goes without saying that the human should be one of the very best.

It seems like there is diminishing interest in these matches, probably because they are considered lost for the humans. But I think in the last two big matches, the players participating (Kramnik and Adams) were wrong choices as far as picking the most dangerous opponent is concerned. So somebody else could still have chances.
I think that if a human/machine match is to be successful for the human, the human will have to spend a lot of time in preparation to discover what strategies work against the program in question.

Most GMs do not use anticomputer tactics against the machines and suffer because of it.