Author: Rolf Tueschen
Date: 05:40:42 02/17/03
Go up one level in this thread
On February 17, 2003 at 08:05:52, Uri Blass wrote: >On February 17, 2003 at 06:57:26, Rolf Tueschen wrote: > >>On February 17, 2003 at 06:08:16, Uri Blass wrote: >> >>>On February 17, 2003 at 05:15:52, Rolf Tueschen wrote: >>> >>>>On February 16, 2003 at 21:14:34, Peter McKenzie wrote: >>>> >>>>>>The point is easy. The whole line, Peter, is bad for Black. See the post from >>>>>>Michael. I gave the number to Peter Berger. Michael showed, and others before on >>>>>>the servers, that White had advantage. Not to be proven with whole lines of 40 >>>>>>plies but still visible. Of course Kasparov played exactly NOT the line that led >>>>>>to advantage. So back to Nxg4. After g4 played, yes, then Nxg4 might be the best >>>>>>move. But NOT in a sense that it's a good move. If Nxg4 is a "good" move then >>>>>>the whole line is bull, that is the summary of that line. And the early 0-0 is >>>>>>the reason for that mess. So, to begin with you must avoid to play 0-0. >>>>> >>>>>But Bob wasn't talking about O-O, he was talking about Nxg4. And I answered >>>>>Bob, so I was also taking about Nxg4. Yes, we were both talking about Nxg4 I >>>>>think. >>>>> >>>>>Get it? >>>>> >>>>>Repeat after me 100 times: we were talking about Nxg4 >>>>> >>>>> >>>>>Having a conversation with you is funny: >>>> >>>> >>>>Since English is Peter's main language the responsibility is _Peters_ that >>>>nobody is to be spoken to like that. Peters indecency is well documented here >>>>[it was just a couple of days ago]. The point here in our debate is clear to >>>>every good chessplayer [Peter is one] and therefore it's telling that exactly >>>>Amir's report is so confusing. And against Peters own knowledge he keeps on >>>>roaring here with a lingual overfloading as if he could defend Amir against Bob. >>>>This message here is a good example for Peters temper and communicative >>>>weaknesses. >>>> >>>>Just to explain to readers with less chess skills I'd like to show why Bob's >>>>position is ok and Amir is wrong. >>>> >>>>The question is, if the move Nxg4 is good, acceptable or bad? >>>> >>>>Amir said: it's the only move, and he insists that Crafty would play the worse >>>>h6 and so that Amir thinks that Junior is better than Crafty because it found >>>>the "best" = only move Nxg4. >>>> >>>>My comment: Amir is fatally wrong! I would say that a program that _avoids_ the >>>>opening of the g-line is _always_ "better"! Although its response (h6) is _also_ >>>>bad. Here it's very clear that such a seemingly contradictory situation cannot >>>>be solved in that move itself but only in the perspective of the whole line. >>>>Every good chessplayer simply goes back and seeks for improvement. And it's very >>>>clear again that already Blacks O-O is very bad because after the response g4 >>>>the threat g5 is already the end. And for basic reasons Black cannot open the >>>>g-file if White has not yet castled. Would Crafty avoid O-O? Michael showed that >>>>it prefered Bb7! So I conclude: Amir confuses the whole question of that line. >>>>He is not even aware of the fact that the whole line is bull if afterwards Black >>>>is forced to open the g-file. Crafty is much wiser because it avoids the whole >>>>zwick! >>>> >>>>Amirs difficulties have a historic record. A predescent of todays Junior was >>>>unable to mate with B+N against the naked K. More, it was incapable to >>>>underpromote. The logic of the response was that these situations were rather >>>>seldom in the games. Perhaps the opening of the g-file is also good in general >>>>and only in very rare cases disadvantageous. Reason: Junior needs the open game. >>>>But here is a clear indication of the fallacies of such lopsided reasoning in >>>>CC. Once such weaknesses become known a human chessplayer will exploit it >>>>without mercy. If several weaknesses exist the overall result will be a real >>>>pain for CC. The difference between comp-comp chess and comp-human chess is >>>>apparent! [With the Weak-Chain theory I had in mind that human players are >>>>capable of aiming exactly to such weaknesses while comps are totally incapable >>>>to _do_ that or to _defend_ against it!] >>> >>>I disagree with the weak chain theory. >>> >>>If I know that my opponent cannot win KBN vs K and cannot underpromotion then I >>>cannot exploit this weakness by getting that endgame. >>>There are weaknesses that are not important for games. >> >>It was just an example. Surely these two weaknesses are not crucial. >>Weak-Chain-Theory means something else. Of course the weaknesses must have a >>basic importance. >> >>My argument was a general weakness in Amirs logic. And with the Nxg4 "only move" >>his logic finally became senseless. Because the open file weakness is letal. >> >>Rolf Tueschen > >The question is if the open file weakness is relevant for that game. > >It is possible that Junior did not plan to open the file by Nxg4 some plies >before that move and only later in the game found that alternatives are worse. > >I also think that chess is not simple and you cannot say that black is losing >only because of an open g file and there are positions when open g file against >the king leads to nothing(see later position in the same game when kasparov >played 25.Rg1-h1) > >I think that you cannot trust human evaluations here because humans including >super GM are not close to be perfect. > >I believe that the evaluation of chess programs may often be better than the >evaluation of GM's because the programmers could test their evaluation function >in games against other computers and get more games than the games that GM's >saw. > >GM's mainly play against GM's and are not used to some good moves that computers >play so their experience can be misleading. > >Uri There is certainly some truth in what you say. There is always a strong point. But IMO you make the wrong conclusions. Analogy: Experts and academics often are in danger due to a specific blind spot. (GM for example are seldom CC experts.) But then to conclude that laymen, amateurs, less educated people had the SAME or even better insight into a topic or field, that is ridiculous. GM might have an imperfect understanding of CC. But in general CC experts have no idea at all about GM chess. What Amir said is simply outrageously false. He argues with his comp analyses or supported by some typically superficial remarks by Kasparov, made to excuse his own mistakes. Kasparov pretended that they led CC back to the science origines. And he is making PR? How does that fit together? And under Kasparov's impression Amir claims super-GM status... Did Amir forget how easily Shredder could have won in Maastricht??? Where every GM sees such a technique in seconds? Bull, that is what it is with a commercial program being super GM! Rolf Tueschen
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.