Author: Robert Hyatt
Date: 11:16:09 07/18/00
Go up one level in this thread
On July 18, 2000 at 02:03:06, blass uri wrote: >On July 17, 2000 at 22:35:18, Dann Corbit wrote: > >>On July 17, 2000 at 14:15:54, Ed Schröder wrote: >>[snip] >>>#2. Computers have only a FEW weaknesses and CCC is full of it. >> >>This is intriguing to me. >> >>Here are the weaknesses that occur to me: >> >>1. Blocked positions are bad for computers >> >>2. Computers have bugs in the opening books. If you discover one before a >>contest, you can have an assured victory. If you discover a bunch of them, you >>can get a bunch of free wins. > >It is not relevant in competitions because the computer can use a different >openining book(the questions is not about the weaknesses of the commercial >version but about the weaknesses of programs that play against humans in order >to get GM norms. >> >>3. Computers are highly repeatable. It is possible to beat them many times >>using the same strategy. For some programs, you can beat them with exactly the >>same sequence repeatedly, but most programmers have repaired this absurdity. >>But I think once you find a winning *technique*, you can use that over and over. > >If the version that they play is different than the commercial version they may >see a different move. > >> >>4. Computers are weak in the early endgame (maybe 10-12 pieces left on the >>board) > >I am not sure about it. >Deep Junior won peter leko in the endgame. > >Uri No... Leko made an ugly blunder, which is not uncommon for humans. I wouldn't say DJ "beat Leko". I would say "Leko beat himself". Which doesn't take a thing away from D Junior. It could also have made mistakes and lost, but it didn't. However, it was probably lost earlier in the game but the opponent overlooked the winning move. Then he even played a blunder later on... probably in time trouble.
This page took 0.01 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.