Author: Terry McCracken
Date: 14:49:42 04/07/02
Go up one level in this thread
On April 07, 2002 at 16:57:14, Uri Blass wrote: >On April 07, 2002 at 14:52:21, Terry McCracken wrote: > >>On April 07, 2002 at 14:41:20, Mike Hood wrote: >> >>>On April 07, 2002 at 14:23:46, Terry McCracken wrote: >>> >>>>On April 07, 2002 at 11:57:18, Robert Hyatt wrote: >>>> >>>>>On April 07, 2002 at 10:41:27, Michael Fuhrmann wrote: >>>>> >>>>>>[D] 1b6/8/8/7p/6k1/6P1/8/6K1 w - - 0 1 >>>>>> >>>>>>Crafty, without tablebases, scores this as about -4.5 for white. Shouldn't the >>>>>>score be 0.0? >>>>> >>>>> >>>>>This excites some sort of bug that I haven't seen. the static evaluation for >>>>>this position is 0.00, but the search produces some sort of position where the >>>>>draw recognition fails... >>>>> >>>>>I'll look at it... >>>> >>>>Don't be too hard on yourself Dr. Hyatt, Fritz 7 although close to accurate >>>>gives an eval of -0.81 pawns and Shredder is a little worse than Crafty at >>>>-4.96, well Chess Tiger 14.0 is "out to lunch" between -12 to -11+ pawns! >>>> >>>>Sorry Christophe! You need to fix this problem before you release Chess Tiger >>>>15.0. >>>> >>>>Terry >>> >>>This is a matter of opinion. I think that today's chess programmers have the >>>right to assume the use of 5-piece tablebases. The time spent tweaking the >>>algorithms for cases like this would be better spent on writing better >>>algorithms for middlegame analysis. >>> >>>But it's the programmer's own decision. Maybe he wants to solve this problem >>>anyway for the sake of his own intellectual achievement. >> >> >>It is my opinion, but I feel _all_ aspects of the game need to be thoroughly >>covered, Opening, Middlegame, and Endgame. >> >>It's a _tough_ job but I feel it needs to be done. >> >>Terry > >Non tablebase position seems to be more important and the general rule says >that the stronger side has chances to win in similiar positions that are not >tablebases positions. > >A program that says by static evaluation without tablebases 0.00 may say 0.00 >also in the following position rhat is not a draw. > >[D]1b6/8/8/7p/7P/6Pk/8/6K1 w - - 0 1 > >Uri Interesting Uri....yes White is "Dead Lost" if 1.Kf1..Ba7!++ if 1.Kh1..Ba7! 2.g4..hxg4 3.h5..g3 4.h6..g2# Fritz 7 sees this instantly! It considers 1.Kf1..Ba7 with nearly a negative 5 pawns. If I were to let it go really deep it might find a long mate, I'm not certain. Terry
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.