Author: Gerd Isenberg
Date: 14:12:32 07/29/03
Go up one level in this thread
On July 29, 2003 at 15:52:36, Ulrich Tuerke wrote: >On July 29, 2003 at 14:19:37, Gerd Isenberg wrote: > >>On July 29, 2003 at 12:11:21, Ulrich Tuerke wrote: >> >>>On July 29, 2003 at 11:59:06, Ulrich Tuerke wrote: >>> >>>>On July 29, 2003 at 11:56:12, Steffen Jakob wrote: >>>> >>>>>Hi Uli! >>>>> >>>>>On July 29, 2003 at 11:50:14, Ulrich Tuerke wrote: >>>>> >>>>>>>[D]8/p7/2K5/1P6/1k6/8/7p/7B w - - 0 67 >>>>>>> >>>>>>>Is there an engine which sees that this is a draw? Here is the complete score: >>>>> >>>>>>Comet sees it (without TB); I guess, many engines will: >>>>> >>>>>At least Hossa and Fritz 8 don't. >>>>> >>>>>>Comet-B.60 Tue Jul 29 17:43:34 2003 >>>>>> >>>>>>depth score sec nodes pv >>>>>> 1. -0.98 0 37 h1e4 >>>>>> 2. -0.98 0 118 h1d5 b4c3 >>>>>> 3. -0.84 0 594 h1d5 b4c3 c6d6 >>>>>> 4. -0.40 0 991 h1d5 b4c3 d5h1 c3d4 >>>>>> 4. -0.40 0 1750 h1d5 b4c3 d5h1 c3d4 >>>>>> 5. -0.67 0 4167 h1d5 b4a5 c6c5 a5a4 d5b7 >>>>>> 5. -0.67 0 7274 h1d5 b4a5 c6c5 a5a4 d5b7 >>>>>> 6. -0.29 0 10668 h1d5 b4c3 c6d6 c3d4 d5c6 d4e3 >>>>>> 6. -0.29 0 20066 h1d5 b4c3 c6d6 c3d4 d5c6 d4e3 >>>>>> 7. -0.08 0 43767 h1d5 b4a4 d5e4 a4a5 c6c5 a7a6 b5b6 >>>>> >>>>>Do I interpret this correctly that Comet sees an advantage for black in the >>>>>first 7 plies? Bishop and pawn vs. 2 bishops? >>>> >>>>Yes, that's right. >>>>What have I done ? >>>> >>>>I guess, I'll have to have a look into the code to understand. -:) >>> >>> >>> >>>Looks like, it's black advanced passer leading to this score. The value of a >>>lonely bishop decreases in Comet's eval with decreasing number of pawns on the >>>board. >>> >>>OTOH, Comet knows that king + bishop alone can never win. So there is an upper >>>limit of 0.0 for these kind of positions from white's point of view. White can't >>>move its king without loosing it's pawn, which is a draw. >> >>Hi Uli, >> >>really great! Is all that knowledge in static eval for this special ending, >>bishop+pawn against pawns? > >Hi Gerd, > >I doubt that it's so much knowledge. I guess, it's just by chance that the >poorly tuned eval could "solve" this. It just thinks that black is better; so >white tries a repetition. -:) I see, the "lucky" heuristic ;-) > >Seems that Uri has detected something where it doesn't work. but sometimes :-( Isi finds white slightly better, due to control of the promote square. I also have some penalty of lonesome bishop with lonesome pawn against pawns, but not enough here. I put this position in my pattern todo list, thanks Steffen. There is really some stuff to generalize in this special ending... Cheers, Gerd > >c/u, >Uli > >> >>Regards, >>Gerd >> >> >>>Black can't move its >>>king much either without loosing the game. It has to have a look on white's >>>pawn. >> >> >>> >>>Thus, we see white bishop and black king moving back and forth, leading to >>>repetitions. >>> >>>Uli >>>> >>>>Uli >>>> >>>> >>>> >>>>> >>>>>Greetings, >>>>>Steffen.
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.