Author: Thorsten Czub
Date: 13:36:36 11/05/99
Go up one level in this thread
On November 05, 1999 at 14:31:45, Laurence Chen wrote: >Does it mean that games played by humans, especially GM's who study openings day >and night, that when they study the games of their opponents and to find out >which openings they play well, and then avoid it completely it's a trap. That's >a very loose definition. A trap in chess is defined as a swindle where the >player falls for a tactical mistake. NO - when somebody prepares a special opening line, against e.g. predecessors of the opponent, he can try out at home how the predecessor would behave. Alexander had played lots of games against chess programs in this opening. But the chessprograms do not know this. human beeing could know this. but how shall a program know: oh god - this time alexander was in charge, i should avoid kings-indian. there is NO tactical thread in the opening. but the trap is that HE KNOWS that programs play this opening weak. I remember Bronstein in a game at the aegon tournament. he had white. the first move was to make. against cstal. bronstein considered. first i thought he is not ready to play. than i thought, strange. he considered half an hour on the first move. the next moves were all in an instant. i was not worried because cstal was in book. than suddenly i saw WHY bronstein considered so long for the first move: he tried to find a position from his experience against human players where HE knows from his experience against chess programs or thought the machine would eat it although it is forbidden to eat the pawn. he considerd which line would fit. which opening he could force the computer. and after he knew it, he easily outplayed all moves. and when cstal considered about eating or not eating and my heart died , he stand up and walked arround having a chat with some friends and was not seen anymore. i am sure he knew so sure that he wasted some time because he knew: if cstal eats the pawn the game is over. this is what remembers me to when i replay the tiger-quest game. alexander knew it would be a good position. somehow like bronstein. the trap is not the tactical topic on board, the trap is that the chess-program has no further experience, has no eyes to see alexander. as a good operator you can watch this. you can invest time to find out about these famous openings . watch out carefullly the oppoents preparation. and change just a second before the game starts the parameters. I always decided which opening lines/setup a second before the game started because the whole time i studied my opponent HUMAN, is he loading a special book. etc. pp. you should never never never play a deep line you would NORMALLY play. or your program NORMALLY plays. > Because chess engines are less likely to >fall into tactical traps then your definition is wrong. the trap is the fact that the program is unable to learn and watch out the team of the opponent and their famous behaviours. > The better word would be >Chess Tiger fell into some HOME PREPARED ANALYSIS. I have seen nimzo team prepared THAT good that nimzo did almost not compute ONE SINGLE move, and with the first move "out of book" it was alike shredder was checkmated. shredder played the whole time the moves the nimzo team expected. imagine this: you try to win a game and you engine runs directly into a line. and each move your program finds after 3 minutes computation you can see on the screen of your opponent as BOOK MOVE. this is ONE result of using autplayers and misuing the games as opening moves . very dangerous. hiarcs had the same problem on a championship (again paderborn) twice ! > BUT if that's true, then >praise should be given to the Quest team for having done the homework. exactly. but the commercial version sold don't have this extra HOMEWORK. So the playing strength of the commercial version can be different. > I believe >that one should not make excuses for failures. nobody makes excuses for mistakes. you don't understand. this is a discussion. you see it as a duel. there is no duel. i am not excusing- i analyse the facts. to learn out of it. and i am not in charge for the decisions that have been made. i am only guest. watching the event. Why do the people in this newsgroup are only interested in duels instead of the CONTENTS of something. For me the game, the things that happened on board, and behind the scene is important. not who is right and who is wrong. and who excuses. and who not. people only want to see show-downs. don't you see that i want to find out WHAT happened and WHY and that i am interested in the discussion ? >Laurence please.
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.