Author: Uri Blass
Date: 05:41:22 04/03/02
Go up one level in this thread
On April 03, 2002 at 06:21:16, Sune Fischer wrote: >On April 03, 2002 at 06:09:26, Uri Blass wrote: > >>On April 03, 2002 at 05:46:13, Sune Fischer wrote: >> >>>On April 03, 2002 at 05:25:19, Uri Blass wrote: >>>>> >>>>>I suspect you are not really talking about static eval the way I do. >>>>>Did you run you program on a position and let it search only 1 node?? >>>> >>>>I do not rememebr thew relevant position now but I remember that in at least one >>>>of my games I believed that had an advantage when I had no advantage and the >>>>computer knew immediatly that I have no advantage(I could not tell it to search >>>>1 node but I could see that even at small depths it does not see an advantage >>>>for me. >>> >>>They extend pretty deeply even at small depths. >>> >>>>I doubt if the moves of Fritz are significant positional errors. >>> >>>Maybe not, but they do reveal that the positional knowledge of programs are >>>basicly infantile. >>>I do not think it is possible to be 2700 elo strong if you have such poor >>>judgement of the position, this is why GMs only use programs to assist the >>>tactical lines, they do not think much of them when it comes to positional >>>understanding. >>> >>>>>Fritz killed me later on the king side, I can't hold the tactics, the best I can >>>>>do is to get fritz to play 20 nonsense moves, but it always finds a tactical >>>>>shot. >>>> >>>>If it finds a tactical shot then it means that maybe the moves are not nonsense. >>>>There is a simple rule in chess that when your position is really bad you have >>>>no opportunity for a tactical shot. >>> >>>True. >>>Fritz position is not bad, but it takes a while for fritz to crack me open, GMs >>>would be much faster at that. >>> >>>>> >>>>>Of cause you can make it so, it is easy go give a higher score for positional >>>>>gain, but programs do not know when it is correct to sac, you said yourself you >>>>>disagreed with the program :) >>>> >>>>Humans also play sometimes wrong positional sacrifices. >>>> >>>>I gave the wrong move only as an example to prove that programs can sacrifice. >>>> >>>>Both humans and machines may be wrong or right in sacrificing but if I give >>>>cases of a right sacrifice you may claim that maybe the program saw everything >>>>in advance so it is not a sacrifice(something that is not always correct). >>> >>>Hehe, well maybe. >>>I have however never seen a program sacrifice a bishop or knight into the >>>opponents king position if it did not see the mate or material gain 10 moves >>>later. >> >>I guess that you did not see the games of chess system tal. >> >>Another point is that I also saw humans sacrifice a bishop or a knight into the >>opponents king position when the sacrifise was wrong and computers do not >>suggest to sacrifice from the first ply and you cannot ignore these cases when >>you compare between the static evaluation of humans and the static evaluation of >>computers. > >The point is, that it is probably correct from a positional point of wiev, but >the tactics may make it wrong. The picture is reversed, humans make tactical >error but computers make positional errors. > >I do not think programs has the postional knowledge of a 1400 player. I do not agree I do not know how can you say that a sacrifice is justified positionally and not tactically. It is justified or it is not justified. A better positional evaluation only means that the evaluation is more often correct. There are humans that are better than 1400 who play unsound sacrifices and win against 1500 players only because their opponents blunder. If a human evaluates that his sacrifices are correct and is wrong in 90% of the cases then I can say that his evaluation is wrong and he will never learn by playing against 1500 players because he wins most of the games. A player can have a rating of more than 1400 without knowledge that bishop and knight are usually better than rook and a pawn(something that programs know) I think that people underestimate computers. I remember that in the match Junior-Gulko in the first game Junior evaluated one of the positions as a draw when I evaluated that Junior has a clear advantage. The relevant position was posted at the time of the first game of Gulko against computers by GCP. I think that Junior7 had better positional evaluation than me and I was too materialistic. I do not know the static evaluation of Junior but I believe that the static evaluation was also close to a draw score. I believe that in previous stage of the same game Gulko had the advantage and Junior did not understand it and it only suggests that there are positions when Junior's evaluation function is better than my evaluation and there are positions when the opposite happens. Uri
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.