Author: blass uri
Date: 12:57:54 06/12/98
Go up one level in this thread
On June 12, 1998 at 10:45:15, Thorsten Czub wrote: >The main problem in the game is, that white plays very simple. >For me the plan is a very very typical plan in the BDG. >I am sure the computers don't understand the plan. >Qd1-d2-f2-h4 is a normal way to develop the queen in this variation. >Also a3 is a standard move. the main problem in the game is that black took the white bishop at g5 if black played another move instead of h6xg5(the simplest is Nd5) I do not see how white can win. The main problem of the program was that it did not analyze enough moves like Qh6 and ideas like 0-0 and after it Rxf6. If it knew it should analyze it(lit knows to analyze captures and checks) then with the same evaluation function It would not lose. >Blacks e6 is the EUWE-defense and is one of the weakest defenses. >Because black helps white to make the game. >a3 is important not to allow the bishop to pin Nc3. White KNOWS that it >needs the Knight to move from c3 to e4 (as is almost always happens in >the BDG). > >What worries me is that computers don't understand what happens. And - >since there is nothing in the search horizont, they see it in the moment >the disaster is not to stop anymore. This is bullshit. >Our computers shall play THERE game. If the brute-forcers want that they >increase their computer-strengths, they should not on the other side >oversee that sometimes a machine would be do a needed job to UNDERSTAND >what the opponent tries. > >I will look for another game of those kind. I don't want to show how A >beats B. >I am not interested in bean counting. I want to show the programmers how >simple today's programs lose. >If grandmasters would understand how weak chessprograms play (normally >the advisors tell them: oh - the machines are horrible strong ! I am >sure Friedel is ONE main reason why Kasparov lost against Deep Blue, >because Friedel told Kasparov how strong the machine is. I doubt this. >These machines sometimes impress people, but i know many situations >where humans can only laugh about their way of playing.). > >Will search for the 2nd example (not with junior, but this time with >genius5.) and post it here, one moment. Have to find it... > >I hope the pattern comes through...
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.