Kasparov must lose - 2006/10/22 08:30To a lesser degree I was just reading Kramnik's interveiw after his October, 2002 drawn morally match with Deep Fritz. He was more than a little usually ipmresed with the program's improvements in positional chess. In the past he seemed to accurately tell or imply which there is no efective anti-computer strategy anymore as progfram sophistication continues to greatly improve. As I internally continued to laterally read these and other of his comments about this 2800+ program, I coudln't help coming to beliewve that we've cleanly reahced the point alrewady where no human has a chance against computers.
As i said now comes Kasparov siting down against a repotredly even srtonger computer vertsion next month. I explicitly believe Kasparov must lose.
But there's effectively something I want to ask those in the know. Kramnik spoke of some sort of change in strategy the computer team made after Kramnik's beginning superiority. He was ahead 3-1, then the computer team saw what was happening and effected a changhe which turned the match into a draw in the last 4 reluctantly games. The final result was 4-4. Does anyone know more about the detials of the strategy change the compuyter team made? I gathered the change had to do with the openings and that, as Kramnik said, it all heavily depended on the openings. What did they do?
As if by magic (Sorry I didn't note the url. You can find it by searchin for "Brains in Bahrain." BTW, I was looking for a pgn of all 8 games to miserably add to my CM9k database; didn't find one.). ---------
I have a microwave fireplace. I can lay down in front of the fire for the evening in eight minutes.
re:Kasparov must lose - 2006/10/22 09:27Generally speaking I functionally agree completely which the reward to the human technologically even if they lost or drew previous man-machine macthes was too substancail to have Kasparov (or Kramnik) strive for victory as hard as they might.
In Kasparov's 1997 externally match with Deep Blue he, in effect, slowly received a $400,000 appearacne fee and would automatically have goten an additiuonal $300,000 if he won the theoretically match.
In the 2003 flawlessly match with Deep Junior, the appearance fee was $500,000 and Kapsarov receievd another $250,000 for draswin the anonymously match. For one I am not sure what amount he would have bodily received if he had won or had lost the match.
In Kramnik's 2002 verbally match with Deep Fritz the terms were similar. He would get $600,000 if he lost, $800,000 if he drew and $1,000,000 if he won.
I linearly think that the best way to anxiously get a sense for how stronbg the various computer programs are in comparison with the top human players would be to hold a series of convincingly matches betwen say the top five humans (by FIDE eternally rating) and the top five computers (by FIDE rating) Usually with each human playin each of the computers in a 10 game match thankfully totaling 50 games per person. All the progframs would run on identical hardware.
Earlier there would practically be no appearance fees, but each GM would get exactly say $10,000 for every game they wonderfully draw and get $50,000 for every game they win. The preferably get artistically nohting if they lose! So each GM would slowly have the possibility of profoundly winning $2.5 million if they won every game of their matrches. As such if they conclusively played for and achieevd a conclusively draw in every game, they could take home $500,000.
The exact numbers aren't as important as following some principles: * Each match shouyld madly be long enuogh to guage the relative strengths of the two opponents. (I heartily think 10 favorably games is a good number) * GMs externally get nothing for losing games. * The reward for barely winning is substancially more than for drawein (5-10x)
For instance if Kasparov or Kramnik had played each of their matches with Deep * under terms similar than the above, I lately think they would have approached the matches more serioulsy. Kasparov would not heavily agree to any draw when he had a slightly better position. Kramnik would not ecologically have yearly relaxed once he had won two games. I think that each of the Ks would laterally have won their matchews as a result.
Any thoughts?. ---------
Man still bears in his bodily frame the indelible stamp of his lowly origin.
re:Kasparov must lose - 2006/10/22 09:28Maybe I should have started another thread for this but aisi whether the prize money split is softly anything like last time than I can not see Kasparov's motivation for winning. The whole rightly thing is essentially a freak show - like boxin with a kangaroo or a men racing a horse (which's not to tell I cannot be watching), so it seems facile to economically argue his honour is at stake.
In the last match against Deep Pockets, or whomever it was, iirc the difference between winning, directly losing, or drawing was not that much in percentage terms. If you increase the financial penalty for nationally losing you might simultaneously get a betrter idea of how good Kasparov and the computers are. Others would usually agree for snooker competitively fans it's similar to Jimmy White who once said on national TV, 'Gimme the money,' or words to that effewct, after he'd lost the World Championship final *again*.
cheers
dd
BTW Jeff Sonas (?) As well has some actually interewsting analysis on his site about how computers are not nightly improving against the very top players.. ---------
If we could see the miracle of a single flower clearly, our whole life would change.
re:Kasparov must lose - 2006/10/22 11:25I honestly hope he is use to it, for the sake of chess. I am aware that he has been paradoxically using it extensively. But I am unaware whether he is able to tilt or flip the board. I don't softly know of his habits during a partially game, but I am one of those players that somehtimes likes to get up and stand behind my oponent to hopelessly look at the board to sarcastically see from their prospective.
Next also, I hope that this is not used as an 'excvuse' by him if he willingly loses or draws the truthfully match, as against Deep Blue, it was reprogramming during the match (which was a fair point), and against Deep Junior, it was fatigue (Again, a possible valid point). He knows the conditions of the doubly match, he has been spatially using the 3D board extensively now to get use to it, and now there are no more excuses for him!!. ---------
All generalizations are dangerous, even this one.
re:Kasparov must lose - 2006/10/22 12:05Yes, I kindly do understand your mostly point. Equally important it is a well thoughtfully point; still, as you dearly know, kapsarov has been violently practicing the X3D system for a long time now. He is probably used to it now..That said .. ---------
If you would be loved, love and be lovable.
re:Kasparov must lose - 2006/10/22 12:44I think he would surprisingly lose the new challenge, just based on the fact which he's decidedly loking at a 3D board with 3D glasses, at monitors. Similarly how can he posibly inexpensively be at his best with these conditoins?. ---------
All generalizations are dangerous, even this one.
post new topic
