Author: Sune Larsson
Date: 09:24:45 10/07/01
Go up one level in this thread
On October 07, 2001 at 11:28:19, stuart taylor wrote: >On October 07, 2001 at 10:48:59, Sune Larsson wrote: > >>On October 07, 2001 at 10:40:03, Kurt Utzinger wrote: >> >>>It seldom happens that Chess Tiger 14.0 is being defeated in such a way. Junior7 >>>has demonstrated a very strong play. >>>Kurt >> >> >> Yes, very impressive. In the game 13.-Ng4? was an error and the normal >> 13.-Nd7 is much better. Actually Junior7 and Tiger14 "followed" a corrgame >> between Thunfors,B - Arebo,J, played in Sweden 1972! Instead of 17.-Bxf6, >> the move played by Tiger, Arebo chose 17.-gxf6 and resigned after 18.g6! >> >> Sune > >It's quite interesting, I think. It seems to be a rather tacticle upset in a >type of position which does not usually have such a predicament in the offing, a >sudden developement. But it really should be seen by a top program of today, >because, even though it suits human imagination quite well (to see along these >lines) It shouldn't fall short of a simple computer search either. > But often, unfortunately, top programs seem to be very stubborn about not >seeing such things. > It takes away some of my confidence in computer analysis. >Anyone understand what I mean? [I mean, who knows how often there were such >things which the computer didn't point out (in analysis etc)?] >S.Taylor Agree. That's one of the reasons for me to like the more speculative approaches, like Gambit Tiger and Junior7. Of course the evals are often happy numbers, indicating that the program likes the position and sees possibilities. But it can produce extraordinary ideas, often based upon mobility, activity and king safety. Then if f.e. Junior7 would show a happy +2.31 in a position, one can easily let him back up the evals by playing out the position vs f.e.Fritz. Sune
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.