Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Crafty modified to Deep Blue - Crafty needs testers to produce outputs

Author: Ulrich Tuerke

Date: 23:49:07 06/18/01

Go up one level in this thread


>...
>grep "Uli" FIDE*.txt
>
>Can't find.

Yes, you're right, i'm a nobody. I'd better shut up and listen when rated
players like you´are giving us some lessons here.

Vincent, Vincent ...

Comet will slay your program again for this at next ipcc. -:)

BTW, I wasn't commenting on Kasparov's play in any way; so I miss the relation
of your remarks below to our discussion here about search algos and harware.

Nevertheless, have a nice day and best regards,
Uli

>
>Definitely stronger chessplayers will be able to tell you some
>facts about the games. The games are big crap.
>
>Biggest shame on Kasparov of course. He played beginner games.
>In computer-human games kasparov managed for a year or 10 to
>get away with that. In the end he always won without doing a thing.
>
>This time in game 6 he was fooled. This was very good for the
>personality of Kasparov i think. Very bad for chess though.
>
>So there is some very hard proof. Toying like a boy of 2200 he
>played some games, focussing on only giving now and then a pawn
>away, missing simple strategic wins which any 2200+ would never
>have missed.
>
>Of course most of those wins were in opening. Only the press conference
>like is at a CB magazine short before the match is showing the
>real truth.
>
>Kasparov played a baby in his opinion. And he was right. Yeah he
>was pretty right. But even a baby can win that position which
>happened in game 6. Kasparov thought he would even not lose from
>a baby in such a position.
>
>WHAT AN ARROGANCE.
>
>But we definitely can analyse the games.
>
>And the games are bad. BAD. BAD. BAD.
>
>Way under my level, AND I'M A BABY CHESS PLAYER IN THE EYES OF KASPAROV.
>
>He can handle 100 in a simulatenously exhibition of mine level
>at the same time, probably not blindfolded, but nearly.
>
>And yes, i'm not even IM yet. I'll get it i guess one day. I'm not
>trying hard, but i'll get IM if i try a bit harder.
>
>the games are way under IM level. With exception of game 6 from Deep
>Blue side, around 6 big errors a game.
>
>Nothing as bad as a strategic error, and kasparov made strategic error
>after strategic error in many games.
>
>Seirawan is excusing him in a way which is just too friendly: "mr
>kasparov is himself never playing this type of positions, so
>that he misses these basic strategic concepts by not playing the
>better move, this might be an explanation for that!"
>
>It's obvious only one man is to blame for all this. I bet the match
>was good for HIS ego. But it's bad for chess in general. Everyone nowadays
>thinks chess is a solved game.
>
>This in a time that a computer
>cannot even drive a car from place A to B,
>without getting major accidents, with one of the
>problems being that it can't process all the information
>between A and B, just because there is too much data.
>
>Even worse, the only way to put my games into the computer is by
>entering it by mouse. Just scanning my paper doesn't work. Scanner
>software is so bad, that it appears i bought for nothing a scanner
>to scan my chess notations.
>
>And i write not so horrible i think. Nowadays most games i write
>down without making writing mistakes.
>
>But i can't scan the notation into the computer, as the OCR programs
>suck just too much.
>
>Speach, we can go on for hours here.
>
>In short, AI is at a very stupid level,  one of the reasons being that
>the processing power of cpu's is still pathetic compared to what we need.
>
>Yet many people think chess is solved.
>
>Solved because kasparov played another 5 pathetic games, from game 1
>kasparov repeatedly has said: "i shouldn't have won that game".
>
>Read: "i played so horrible that game and i even won it,
>after winning that game i thought i played an unborn baby and the
>next 5 games i thought i was playing one".
>
>You can say a lot, but a few things we CAN conclude:
>  a) the moves played by Deep Blue were very bad
>  b) the good moves played by Deep Blue are also played by todays programs
>  c) nearly none of the bad moves of deep blue are played by todays programs
>  d) Amazingly Kasparov has hardly had critics about his bad performance
>     against deep blue. MOST likely that's because of some lies he said
>     in front of cameras and accusations which simply aren't true. No
>     deep blue team didn't cheat with karpov in the background or behind
>     some kind of monitor influencing the game. All lies from kasparov which
>     reveal the truth. I'm blaming kasparov for something, but amazingly
>     in nearly all articles written about kasparov-deep blue this hardly
>     is getting written down. Only in the chess scene itself this is getting
>     written down
>
>     If my memory doesn't let me down:
>  e) in 1790 there was a machine that played very good chess, so people said.
>     It beated strong chess player after strong chessplayer.
>  f) in 1840 still several people believed that around 1790 a machine
>     was constructed which played very good chess
>  g) in 1900 people knew that there was no such machine, only a machine
>     with a hidden GM inside, so the machine wasn't playing very good chess
>  h) in 1914 there finally was a machine which could play a bit of chess itself
>     of course very limited chess and very bad. WW1 of course overshadowed
>     the introduction of this machine.
>  i) in 1997 we had a computer which beated kasparov in a match, so
>     it was believed, by most people, that it really played very
>     strong chess
>  j) in 20xx it was obvious that the thing wasn't so good
>  k) in 2yyy there really was a strong machine
>
>my hope is that xx gets replaced by 01 when Kramnik annihilates Fritz
>end of this year.
>
>yy is unknown of course. Historically spoken i should fill in 110 there,
>but my own guess is 066.
>
>Best regards,
>Vincent
>
>>>
>>
>>>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.