Author: Charles Milton Ling
Date: 12:02:52 10/09/02
Go up one level in this thread
On October 09, 2002 at 14:34:09, Robert Hyatt wrote: >On October 09, 2002 at 12:31:29, Roy Eassa wrote: > >>On October 09, 2002 at 12:07:38, Kurt Utzinger wrote: >> >>>On October 09, 2002 at 11:06:46, Chessfun wrote: >>> >>>>Point was; >>>> >>>>"The Fritz team has the consolation of having outplayed Kramnik in the opening >>>>stages with the Scotch Opening" >>>> >>>>It never looked that way to me. >>>> >>>>Sarah. >>> >>>I fully agree. It's simply not true that Deep Fritz had outplayed Kramnik in the >>>Scotch opening. >>>Kurt >> >> >>Right. But Fritz did apparently have a perfectly acceptable (approx. equal) >>position until playing 19.a3?, after which Kramnik (amazingly!) thought Black >>was winning. > > >"equal" between two strong humans. But Fritz has an endgame problem, and that >opening >played right into the jaws of that problem. Kramnik knows what he is doing as >he has had >ample time to discover these "problems". And once you know that the program >loses a >bunch of "skill" when queens are removed, then suddenly you can make a few >positional >concessions to dump the queens, knowing that your opponent is going to make >_worse_ >positional concessions in a part of the game it doesn't understand very well >(the endgame). > >The programs do well against other programs with these huge passed pawn scores >and without >knowing much about majorities, candidates, and the like. But _not_ against a >top GM... Exactly, Bob. As some wise person (I don't remember who) said: "It is far better to get a slightly worse position the computer doesn't understand than a slightly better position the computer does understand." Although I am quite willing to accept that the position was balanced for quite a while, I knew in my bones that Kramnik (provided he was playing at his peak - which is beyond praise) would slowly but surely make DF look pretty silly. Charley
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.