Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: programmer's position

Author: Will Singleton

Date: 12:04:53 04/16/03

Go up one level in this thread


On April 16, 2003 at 14:20:27, Robert Hyatt wrote:

>On April 15, 2003 at 23:28:47, Will Singleton wrote:
>
>>On April 15, 2003 at 17:40:17, Vincent Diepeveen wrote:
>>
>>>On April 15, 2003 at 17:23:09, Will Singleton wrote:
>>>
>>>Will,
>>>
>>>Some programs have a very simple qsearch and simply recapture at all squares
>>>where already in the normal search has been captured. Simple, easy and very
>>>fast.
>>>
>>
>>I tried that method by itself (replacing the SEE), and got a good speed boost
>>along with some weird pv's.  However, using (SEE || any pv captures) does solve
>>this particular problem faster.  Not sure if the additional nodes justify normal
>>use.
>>
>>Will
>
>
>how long is it taking you to see that Ne5 is ok for black?  IE I spot this at
>depth=11
>after a couple of seconds.  At depth=10 Ne5 looks good for white.

The position looks about equal, except for the aforementioned hiccup.  This is
with the new code, which improves but doesn't eliminate the problem.

8     9   144  586052 Ne5 Qb4 a5 Bb6 Qd6 Bxa5 Qxb4 cxb4 Rxa5 bxa5 Nxc4
9   223   391 1542971 Ne5 Qb4 Nxc4 Qxc4 Rd5 Nb3 Rh5 f4
10  -14   753 3096159 Ne5 Qb4 Ng4 g3 a5 Bb6 Qd6 Bxa5 Qxb4 Bxf7 Nxf7 cxb4
11    5  1813 7592375 Ne5 Qb4 Ng4 g3 Nxc4 Qxc4 Ne5 Nb5 Qb7 Qe2 Nf3 Kh1 Bh3

Will



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.