Author: Uri Blass
Date: 03:19:55 10/24/02
Go up one level in this thread
On October 24, 2002 at 06:12:35, Vincent Diepeveen wrote: >On October 23, 2002 at 18:04:12, Robert Hyatt wrote: > >>On October 23, 2002 at 16:18:11, Uri Blass wrote: >> >>>On October 23, 2002 at 15:19:03, Robert Hyatt wrote: >>> >>>>On October 23, 2002 at 15:16:39, Robert Hyatt wrote: >>>> >>>>A couple of classics come to mind that are easy to find. >>>> >>>>1. Kasparov blundered in game 2 of the DB match in 1997. He resigned a drawn >>>>position. >>>> >>>>2. Kasparov blundered in game 6 and blew an opening sequence of moves and lost. >>>> >>>>It is interesting that we see _more_ mistakes against computers. >>>> >>>>My theory is that the computers simply notice the mistakes that humans might >>>>well overlook... >>> >>>Today games of GM's are analyzed by chess programs so this excuse cannot explain >>>finding more blunders of GM's against computers. >>> >>>I also do not remember big blunders of GM smirin against computers and smirin >>>played at faster time control. >>> >>>Uri >> >> >>We were not just talking about GM vs computers. But GMs vs _anything_. The >>point was >>that against humans, many mistakes get overlooked. Against computers, it is >>much less likely >>to be missed unless it is way too deep... > >Your point is wrong. Please show me kasparov-kramnik games >and how many 1b blunders they make there. > >Of course if they have a won position and win the position (even if >there is a mate in 11 or whatever) that doesn't count. > >It's about blowing won positions to a draw or a loss and a drawn position >to a loss. > >That's the key thing. The key is not only tactical mitsakes but tactical mistakes that programs can find even based on search of depth of 1 or 2. A tactical mistake that programs understand only after 7 or 8 plies can be also discovered in less then one second but it is not the same as the tactical mistake that kramnik did in his first loss. 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.