Author: Uri Blass
Date: 02:08:31 01/07/01
Go up one level in this thread
On January 06, 2001 at 14:42:45, David Rasmussen wrote: >On January 06, 2001 at 12:37:03, David Rasmussen wrote: > >>On January 06, 2001 at 10:14:06, Uri Blass wrote: >>>> >>>>Chezzz locks on Ne2 after 44 secs on my machine, and it never considers g3. >>>> >>>>It hasn't found an especially good score, though. -0.20 for white, so I will let >>>>it search for a full 10 hours or so. >>> >>>Thanks >>> >>>I do not say that Ne2 is winning but only that it is probably the only not >>>losing move so there is no problem with score of -0.2. >>> >>>I believe that DB1 could draw the game instead of losing with Ne2 instead of g3. >>> >> >>Chezzz has now searched for 3 hours, and still considers Ne2 best with a score >>of -0.25 at depth 15 (Chezzz has inferior branching factors, so better engines >>will search deeper in this time). g3 is not considered as one of the first moves >>at the root, it is sorted down as move number 6 or so, so there are 4 other >>moves that Chezzz would try after Ne2 but before g3. >> >>I will try to run it after g3, and see what it finds. > >I've now run it to depth 14 where the output is this: > >14 +0.45 1... Qg5 2. Ne2 Rxd2 3. Qxd2 fxg3 4. Qxg5 gxh2+ > 5. Kxh2 hxg5 6. Kg3 Be6 7. Nc3 Bf5 > 8. Nb5 a5 9. Nd4 g6 (506539 kN) > >So, the value is 0.45 whereas the value for Ne2 was 0.25 for black. That's not a >big difference, but at least it never considers g3 seriously. I can't imagine >why DB would play that. Lousy eval, maybe. Byt lousier than Chezzz?? I don't >think so... It is possible Hiarcs7.32 has also worse evaluation than your program. It considers g3 as best at the end of iteration 11 with score of -0.70 but I guess that it is going to see bigger problems with g3 and change its mind to Ne2. My experience with chess programs suggest that all of them are going to see Ne2 if you give them enough time(I guess that some programs give big bonus for the passed pawn at e3 after Ne2 Rxd2 Qxd2 Qa1+ Nc1 Qe5 Ne2 e3 fxe3 fxe3 and I believe that they are wrong). 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.