Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: developing Junior (and other pro programs)

Author: Vincent Diepeveen

Date: 06:36:48 09/01/02

Go up one level in this thread


On September 01, 2002 at 07:07:33, Georg v. Zimmermann wrote:

Yes i very well realized how arrogant the statements from Uri were
in direction of Frans and other programmers.
He isn't capable of realizing that Frans is nullmoving even if it
is improving his node counts 0.01%.

I didn't do arrogant statements other than in the direction of Uri,
who is writing a lot of things here down like a kid. I wonder why
you replied onto my posting as if i have written the text, quoted
by you, which in fact Uri has written.

If Frans can optimize his engine from 5000 instructions a node to
on average 4999 instructions a node, then Frans WILL DO IT.

Even if it takes 1 month of fulltime work.

In DIEP i don't nullmove if beta >= mate-1000, that doesn't hurt
double nullmove properties, and it prevents me from losing nodes
when figuring out PV.

Uri probably hadn't thought of is that in MTD you don't have ever a
score >= mate-1000. In fact you only have a bound. So if you don't
care for the true value of it, then you can always nullmove after
nullmove.

>
>>Obviously you need to use nullmove first move after pv always
>>when you don't use alphabeta, but MTD instead.
>>
>>>I remember from analyzing another study that
>>>this kind of behaviour is in Fritz for years
>>>(Fritz3,Fritz4,Fritz5)
>>>and I wonder why Fritz insist to use null move in
>>>the wrong way
>
>Do you two realize how arrogant those two statements are ? Especially Diepeveen
>has a strong program, but as long as Fritz is years ahead, I would not make such
>statements.
>I bet Morsch knows more about NullMove than we 3 together.
>
>Regards,
>Georg v. Zimmermann



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.