Author: Uri Blass
Date: 05:35:13 08/28/03
Go up one level in this thread
On August 28, 2003 at 08:22:49, Mike Hood wrote: >On August 28, 2003 at 08:04:03, Mike Hood wrote: > >>On August 27, 2003 at 23:34:38, Rex wrote: >> >>>Have fun with this one... Let your engine run for a few minutes. Interesting >>>to see the different evals on this one.. >>> >>>[D]3r2k1/1pp2ppp/3r4/1P1npq2/3n4/3PPP2/1BQ1BP1P/2RR3K w - - 1 1 >> >>Very interesting... on my 733 MHz Pentium Shredder 7.04 only needs 2 minutes to >>realise that Black is winning. On the other hand, Fritz 8 fails miserably. After >>60 minutes Fritz 8 still evaluates the position as 1.04 in White's favour. > >Oops..... maybe I judged the position too soon. I just let Shredder loose on the >position again, longer this time, and he changes his evaluation back to (almost) >even after 4½ minutes. This is the analysis after 15 minutes. The big jumps are >at 2:49 (0.20 to -7.86) and at 4:21 (-7.50 to -0.40). This is obvious that the only evaluation that you can trust is when an iteration is finished. Shredder saw that one move is losing and changed its mind. Nothing special. It may be more interesting if it is possible to compose a position when the evaluation of top programs are something like this at the end of the iterations: +2.45/1 -1.54/3 +1.65/5 -1.43/8 +1.34/10 -2.45/13 +1.32/15 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.