Author: ALI MIRAFZALI
Date: 14:45:23 02/09/03
From a Computer chess point of view it is very difficult to explain the move 17....Rae8? from Game 1.Although there is no doubt that white stands better; however most computer evaluations go from minus one pawn to minus 2 full pawns after 17....Rae8 is played.This is a one pawn difference ;a significant amount for computer programs that often worry about 1/10 pawns advantage or disadvantages!!! No computer program to my knowlegde plays 17...Rae8? Most programs want to play 17...Rab8 although Kasparov might have still one won the game ;he would have to overcome considerable resistance.But 17...Rae8? is suicide.Enough said for game 1.In game 3 Kasaprov finally loses;but unlike what many claim the 32.Rh5? error is not due to time pressure;after all he is the same person that beat Fritz 3 on a pentium 90 at BLITZ in 1994!! This game shows us the human side of chess ;after all games between human GMs are mostly also decided on the tactical mistakes of one of the participants also.In game 5 Deep junior simply blunders with 10.... Bxh2? How can amachine play a move like this?? Kasparov should have played 16.g3 and I think he would have won ;but he didnot for whatever reason .A good way to test this would be to see if a human GM can hold the black side while he is playing a program that played 16.g3. In game 6 another strange thing happened: On move 23 White plays 23.Nb3? In fact no other commercial program plays this because if one feeds 23.Nb3? then immediately 23...Rxc3!! pops on the screen with an advantage to black.Deep Junior did not see this?! The Conclusions of this Match are 1.Despite world computer championship titles and rating lists there are many programs that are better than DeepJunior. 2.Kasparov only drew the Match which means a beeter program would have beaten him.
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.