Human victory

[jim]

What about a match between five top grandmasters, playing in tandem, with a captain, and the latest improved Rybka. Rybka has at its beck and call ---thousands of masters helping it and millions of master games. Maybe some corporate sponsers would offer some very big prizes. If these five guys could kick rybkas butt---it would be a human victory wouldn't it?

[Dann Corbit]

What about a match between five top grandmasters, playing in tandem, with a captain, and the latest improved Rybka. Rybka has at its beck and call ---thousands of masters helping it and millions of master games. Maybe some corporate sponsers would offer some very big prizes. If these five guys could kick rybkas butt---it would be a human victory wouldn't it?

I doubt that 5 top GMs working in tandem would do better than one, if the one was Kasparov or Anand or one of the elite.

I would like to see correspondence chess matches with top correspondence GMs and top computer programs. When it comes to human/machine matches, that is the only thing left that keenly interests me.

There (I think) the humans still have a chance of victory for a match.

[Ryan Benitez]

Given long enough time control the correspondence players would beat the top computer program very easy. No engine I know of is close to closing this gap at time controls of greater than 24 hours per move.

[geots]

What about a match between five top grandmasters, playing in tandem, with a captain, and the latest improved Rybka. Rybka has at its beck and call ---thousands of masters helping it and millions of master games. Maybe some corporate sponsers would offer some very big prizes. If these five guys could kick rybkas butt---it would be a human victory wouldn't it?

I doubt that 5 top GMs working in tandem would do better than one, if the one was Kasparov or Anand or one of the elite.

I would like to see correspondence chess matches with top correspondence GMs and top computer programs. When it comes to human/machine matches, that is the only thing left that keenly interests me.

There (I think) the humans still have a chance of victory for a match.

I remember in one of Fishers games the 3 grandmasters who were annotating the game in another room for the audience declared that Fischer had a lost game just at the moment his opponent was resigning. So more is not necessarily better. Especially with differing opinions while on a time clock. I imagine it might be worse than just 1- if the 1 was as Dann said. At any rate, welcome to CCC, James. Newbies get hassled a bit from time to time- if that happens just dont pay any attention. Its a rite of passage of sorts. But it's a nice forum. Enjoy it.


Regards,

[Dann Corbit]

Given long enough time control the correspondence players would beat the top computer program very easy. No engine I know of is close to closing this gap at time controls of greater than 24 hours per move.

It's not always so easy, even with great correspondence players:
http://correspondencechess.com/campbell/ham/ham.htm

Nickel did win his match:
http://www.chessgames.com/perl/chess.pl ... pid2=87303

I guess that Rybka on a giant pile of SMP CPUs might be a handful for the world champion, even at one move per day or so.

[Uri Blass]

Given long enough time control the correspondence players would beat the top computer program very easy. No engine I know of is close to closing this gap at time controls of greater than 24 hours per move.

I do not agree.

Even top correspondence players when they use a computer to help them cannot always win even against relatively weak program like Fritz10.

See the following game when it seems that GM Schön Wolfram failed to win against Fritz10 and this player has an iccf correspondence rating of 2696

http://ancients.correspondencechess.com ... opic=147.0

http://www.iccf-webchess.com/PlayerDeta ... x?id=81128

Edit:Note that the GM did not have Fritz10 but in a real match of computer against human the human can also not have private rybka or private zappa or at least not have the special personality that is going to be used for the match.

Uri