Author: Robert Hyatt
Date: 11:36:16 10/26/02
Go up one level in this thread
On October 26, 2002 at 07:47:20, Rolf Tueschen wrote: >On October 26, 2002 at 03:12:45, Uri Blass wrote: > >>On October 25, 2002 at 22:36:56, Robert Hyatt wrote: >> >>>On October 25, 2002 at 18:22:29, Rolf Tueschen wrote: >>> >>>>On October 25, 2002 at 12:18:38, Robert Hyatt wrote: >>>> >>>>>On October 25, 2002 at 09:55:48, Rolf Tueschen wrote: >>>>> >>>>>>On October 24, 2002 at 17:17:19, Robert Hyatt wrote: >>>>>> >>>>>>>Here is a game where Kramnik played black, and he made the same sort of >>>>>>>sacrifice he made against Deep Fritz, >>>>>> >>>>>>Wrong! >>>>>> >>>>>> >>>>>>>but he made it against Anand, and he >>>>>>>got rapped for it. Crafty says this just drops a piece. I ran the position >>>>>>>after >>>>>>>Bxf2 to a pretty deep depth and the score didn't change as I even followed the >>>>>>>game for a move or two deeper. If you let Crafty search for a black move >>>>>>>there, it likes either the rook move as suggested in the annotation, or Nd8 >>>>>>>if you let it search longer than one second... Score after Bxf2 is +3.5, score >>>>>>>after Nd8 is +.5. >>>>>> >>>>>>I think we could prove very quickly why it's way too early that we could rely on >>>>>>the research with our actual PC programs. Also you miss the complete problem of >>>>>>that game. It's not at all the same or similar to the Nxf7 against Deep Fritz. >>>>> >>>>>It is similar for one main reason: It simply drops a piece for nothing. >>>>> >>>>>So in that regard, it was a sacrifice that was unsound, which could be called >>>>>a blunder by most any definition since it causes an equal position to turn into >>>>>a dead lost position. >>>> >>>>What are you talking about? The position IS already lost! So it is NOT similar >>>>to the Kramnik-Deep Fritz position! >>>> >>>>Rolf Tueschen >>>> >>> >>>I'm not sure what you are talking about. The position before the sac is not >>>lost for either side. >> >>I did not analyze the position but it is clear that one of the following is >>correct: >> >>1)The position is not lost for either side >>2)The position is lost for kramnik but Crafty does not understand it. >> >>If 2 is correct then it means that it is not similiar to the kramnik-Deep Fritz >>blunder and the sacrifice may be a desperate attempt of kramnik when he found >>that other lines are bad. >> >>Uri > >And what do you think is the case? <g> > >I said it in my first statement in the thread that Bob could not hope for wise >decisions from Crafty. The move Nd8 instead of Bxf2 wouldn't change much. So of >course 2) is right. It's beyond my understanding why Bob takes Crafty as an >arbiter in this case. I'm not taking crafty as anything in this case. It found the position. But _I_ looked at it for a good while. I won't begin to try to follow the tactics after the sac, because there is no way to predict how the GMs might vary. But the sac seems to be unsound in using a computer to try a few of the variations. Not going for the sac does _not_ seem to be losing. To my eye. Or to Crafty's "eye". My conclusion? Play on without resorting to tossing the knight. > NB that I took this position only because Bob spoke of a >first example for etc. I thought let's take a look and found Anand's own >analysis. 15. h3 was the novelty of his second (GM (!!) I forgot his name!) But >it's strange to see how Bob did not react on the details I wrote. As if I - as >simply Rolf - had made up my mind to challenge Bob's Crafty.. LOL I didn't react because I didn't see anything to react to. The sac was a blunder, because it lost the game. My point from the beginning. Whether he would have lost without the sac is another point altogether. And so far, I don't see any _forcing_ line that loses... The sac made the loss easy. > >No, I would never try that. So, if I cough up something you can be sure that I >have references. And it's simply a fact that after 15. h3 there is no move where >Kramnik could have saved the game - in terms of GM chess, of course Crafty might >think different for a looong while. But all the PC progs are in the same >difficulty I suppose. > >BTW the whole discussion is also a good example for the nonsense with GM books >and my claim of fair conditions after FIDE rules. > >1. There is not even time enough to let all these "questions" in millions of >variations be tested by the progs. > >2. The progs don't have the class to either understand the bad sides for Black >in the Anand-Kramnik game, after h3, and they are clueless of what to do to >evitate the mess. Like humans who throw the whole variation with Na5 into the >trash. I disagree. The programs are quite capable of following _that_ kind of position. It isn't particularly a positional struggle, and that plays into the strength of the machines. Not that they are perfect in tactical positions always, but they are _very_ good. > >3. The cheat would exactly start here if say one Alex Kure decided to tell >Fritz that "this Na5 line is bad", because the prog in real might find it >playable. So Kure would play the boss in the understanding of Ingo Althöfer. But >I always read that Fritz etc are GM _themselves_... Which is hyperbole to say >the least. And Bob says the same. > >I will keep it short for now, in awaiting also Bob's arguments. > >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.