Author: Rolf Tueschen
Date: 09:40:12 05/05/05
Go up one level in this thread
On May 04, 2005 at 23:04:30, Robert Hyatt wrote: >On May 04, 2005 at 20:01:39, Rolf Tueschen wrote: > >>On April 30, 2005 at 15:34:28, Robert Hyatt wrote: >> >>>So I don't buy his "I didn't think the machine would make a mistake." nonsense >>>because it implies he knows something that he couldn't possibly know. I've seen >>>_many_ games on ICC where Crafty was a piece ahead but forced to take a >>>perpetual to avoid getting killed... >> >> >>Err - sorry, but I don't quite understand your speech here. You're saying that >> >>- you was a piece ahead, so with +2 at least on your screen? >> >>- then YOU were forced to TAKE (?) - you mean to allow? - a perpetual? > > >No. What I saw was an evaluation of 0.00, but counting material myself, crafty >was actually ahead. Ok, then all the questions I had are answered. I thought you were talking about a discrepance between the shown value on the display (here for example +3.00 and the practice of accepting a draw by the machine...). I simply misunderstood the English. Thanks for the clarification. > I have seen it as far ahead as +5 if you count material, >but the score is 0.00. I have seen it -9 in material, with a 0.00 score. That >is, it might be a pawn down, then sacs the queen to open the king to a forced >perpetual... > >counting material alone means nothing, without knowing what the search is seeing >to go along with that material count... Of course. Here they had a good example with the imprisoned Rook and every machine will show +3... for itself although that R has no future. That was an example for Kasparov for the general inabilities if a machine - thus shown in one of the ChessBase Mags before the match in 1997. At the time it was clear that Kasparov couldn't adapt to the reality that such stupid machines should be able to give him a real hard time. Out of experience we can expect Kasparov having such difficulties. On the other hand it is clear that the continual calculating requires a superb physical stamina; but also here a Kasparov was handicapped because of his superiority in human chess. It's difficult to admit and then train for the reality that a generally stupid machine should cause you so many difficulties in practice. You certainly never had such thoughts but here we must consider that such super GM have an elitarian socialisation in human chess and NOT everyone is able to really do the very hard work against a machine. Not because of the genius of the machine but its constant flow of faultless play up to a certain depth. That must be a tiring experience. I don't know all the details, but in human chess this is why Lasker was World Champion for decades, why Capablanca was such a successful winner at chess and also why a machine can be a terribly dangerous terminator against human players who don't succeed with their superior chess. In my view Kasparov had such a difficult task because his preperation was completely wrong. Perhaps he isn't the best human player anyway against a machine because his whole tricks are useless which are so dangerous for human opponents. But anyway this is no reason to think lightly about Kasparov. If you castrate his best talents by holding the opponent in secrecy then you can't expect that a Kasparov could show his class. All matches after the one in 1997 already changed the odds simply by rules for the "allowed secrecy" of the machine. Most shows ended in a final draw. To the best of the financial needs of the GM. > > > > >> >>- because (??) you SAW that if you played on your advantage (visible trough the >>+2) would in the end mean losing the game??????????? > > >Yes. > >Imagine a position where I sac my queen for an apparent forced mate (I am a >human). Early searches will show you with a score of +9 (you are a queen ahead >since I just gave you mine.) but suddenly, now, as the search goes deep enough, >you see you are getting mated, but you spot a perpetual I didn't see. And you >take it. You are a queen ahead, but you take the perpetual. Why? Because even >though you are a queen ahead, you are really about to lose your king... > >> >> >>And you are talking about CRAFTY who's seeing this???? > >Yes, except that crafty would correctly show a 0.00 score since it sees that it >must go for the draw. I can remember ACM events where Cray Blitz wanted to >resign, but Mike Valvo would say "no... Bob, look at the board." And when I >did, I noticed material was equal. His point "spectators might not see the >tactical problems and consider this equal." > >This is pretty common. > > >> >>Excuse me, but I'm not believing in the supernatural of THAT idea... >> >>BTW this time I will take this through to the end. I'm not interested in draws, >>this argument I will transform into a clear win! ;)
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.