Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: How is Hydra faster and better than Deep Blue?

Author: Robert Hyatt

Date: 22:11:41 05/30/05

Go up one level in this thread


On May 30, 2005 at 17:55:57, Ricardo Gibert wrote:

>On May 30, 2005 at 13:37:55, Gian-Carlo Pascutto wrote:
>
>>On May 30, 2005 at 11:27:46, Ricardo Gibert wrote:
>>>>>But which also reduces EBF. You should have left out this paragraph.
>>>>
>>>>SE definitely INCREASES the EBF.
>>>>
>>>>--
>>>>GCP
>>>
>>>Generally, even bad extensions reduce the EBF.
>>
>>Why would they? That makes no sense.
>
>Extentions are a form of selective search. Why shouldn't they?

They really aren't.  They just extend a branch here and there deeper than
normal.  But other branches are still searched just the same...

Now if you want to argue that the "extended depth" is the normal depth, and all
the non-extended branches are searched to a lesser depth then perhaps that case
could be made.  But for SE, the problem is that it isn't just an extension that
gets triggered based on some static characteristic like whether or not the king
is in check.  It is necessary to do a significant search (albiet to a shallower
than normal depth) to determine whether the extension should be triggered or
not.  And that is pure overhead increasing the EBF significantly.


>
>>
>>>The only way I can think that could happen is if some type of redundant searching
>>>is going on.
>>
>>SE needs a lot of extra searching to establish singularity.
>>
>>>Thank you for reinforcing my disinterest in SE.
>>
>>It's possible to do better than the DB implementation/method.
>>
>>--
>>GCP



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.