Author: Robert Hyatt
Date: 16:52:26 11/11/98
Go up one level in this thread
On November 11, 1998 at 04:25:06, Reynolds Takata wrote: >On November 10, 1998 at 17:19:21, Robert Hyatt wrote: > >>On November 10, 1998 at 10:29:12, Reynolds Takata wrote: >> >>>On November 10, 1998 at 03:47:08, odell hall wrote: >>> No offense but i'm seriously doubting that those words have echoed in your >>>ears, because Kasparov to my recollection said "if deepblue were to start >>>playing TOURNAMENT chess, I personaly guarantee I will tear it to shreds". >>>Would Kasparov tear it to shreds? Well who knows that's only conjecture. >>>However, just as Kasparov may be drawing too big of conclusions from the >>>match,you are as well. Especially considering the short nature of the match. >>>Heck if Jan Timman had beaten Kasparov ina match or even Judit Polgar had beaten >>>Polgar in a match, nobody and i mean nobody, would be claiming that either of >>>those players were better than Kasparov. Another thing is that you are >>>overlooking the POSSIBILITY of LUCK. I say this, because as a master i know >>>that there is luck. An example, though i am only an average master, in the game >>>that Kasparov resigned that was a draw. I saw the draw almost instantaneously, >>>maybe a minute to check to make sure of it is all. As for everyone not seeing >>>the move that's not true. >> >>I'll bet you didn't see the real draw... > >Well Bob you can bet all you want have a coke and a smile. > >because it is 30 moves deep. And it >>took a bunch of people all night with computers and brains in gear to find this >>. You might have "thought" it was a draw... But that's a long way from >>"knowing" it is a draw. And I'm not trying to be insulting... but a bunch of >>GM players didn't see it either... It's non-trivial... > > > >> >> >>> In fact many players believed there was something. I >>>wont forget I.M. Ashleys shock when it happened he says at the moment something >>>to the effect "What? Kasparov is resigning?" In a very shocked voice. At that >>>point many people just stopped examining the position. Further, 2 more of the >>>games Kasparov was had winning positions(and blew) that most average GM's would >>>have won, >> >>This is not a convincing argument. > >The above statement isn't an arguement. As for them being Winning positions you >need to try to understand what is generally meant when players and Chess >periodicals say winning. You can have mate in two and not win. The above is >not only my analysis but MANY Grandmasters as well, you only have to look around >for the confirmation of this, which include statements by Anand and Orlov. > My point is that two things are happening here... 1. some GM players have a lot of disdain for computers, having been exposed to micros and finding them relatively unintelligent, excepting for fast time controls. And they assume that *they* can analyze more deeply and more accurately. (more on this in a minute). 2. these same players look at some positions and say "white is winning..." and that's that. Care to look thru a good opening book and let's do the following: I can pick *any* opening where the analysis is +- and you have to play the white side of it, with a wager of $10,000 on the outcome. You know what would happen there, correct? Because there are lots of such annotated games where something *serious* was overlooked, and instead of being +- it should be -+. The thing that most bothers me about the comments about game two of the DB match is that *everyone* agreed that the computer should have played Qb6 and won a pawn, with a probable win as the result. And they all agreed that a very strong human might have tried Be4 instead. And *all* overlooked that Qb6 leads to a draw. So what is convincing when they say that in game X, DB was lost, but it managed to swindle a draw because Kasparov made a mistake. Perhaps it was the other way around. The game was a draw due to some deep but unseen tactics, and kasparov played perfectly to avoid letting it turn into a loss. I've been around too many computer chess matches. 20 years ago the masters and IM's were pointing out mistakes here and there in every game. 10 years ago they were pointing out mistakes and occasionally finding they were wrong. At the last few computer events I attended, when the computer would make what appeared to be a blunder, the IM/GM players would start trying to figure out what was going on, assuming that the move wasn't a blunder but was actually forced. A 180 degree turnaround in thought process, because the computers were suddenly right way more than they were wrong. I've seen nothing to convince me that this couldn't have happened in the games that DB drew... ie they were drawn all along, rather than being won by Kasparov until he made a fatal blunder (or blunders) in each... So just because a GM publishes analysis that says that this game should have been won, is *not* convincing. IE check out the comments by GM Shirov. He published some positions that he claimed would take days, weeks or months for a computer to find... and Crafty found most of them inside a tournament time control. Given that level of misunderstanding, how badly do you think they might be underestimating Deep Blue, which *really* understands tactics? > > It might well be that you *thought* those >>positions were won when they were really forced draws. DB was *not* a pushover. >>Without someone showing a demonstrated win I would remain skeptical... > >Bob what else can one expect from a skeptic? That's part of being a scientist. I believe what I can prove. I try to prove what I want to believe. And proving means producing some real data or results... ie playing the game. The most common statement I heard after game 6 was the famous "any micro program could have won game 6, after Kasparov allowed Nxe6." We now know that this not only was wrong, it was grossly. Several have tried this from the white side and no micro has been able to win yet. So I accept this as proof that the earlier statements were wrong. Ie it would be fun to play a couple of GM players starting from that position and giving them white, just to see if they can produce what they say they can...
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.