Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Do programmers mean to the same thing when they say nodes?

Author: Bas Hamstra

Date: 11:40:00 09/06/99

Go up one level in this thread


On September 06, 1999 at 10:51:55, Andrew Williams wrote:

>On September 06, 1999 at 10:09:57, Bas Hamstra wrote:
>
>>On September 06, 1999 at 07:41:01, blass uri wrote:
>>
>>>On September 06, 1999 at 07:05:03, blass uri wrote:
>>>
>>>>I know that nodes in some programs(like Junior) include illegal moves and my
>>>>question is if the same illegal moves are defined as nodes by all the programs.
>>>
>>>I am not sure about the definition of nodes about Junior but I am sure that
>>>Junior generate illegal moves and it discovers that they are illegal only by
>>>search so I guess that it does not count only legal moves as nodes.
>>>
>>
>>Another thing is: the natural way to count nodes is to put a nodes++ in the
>>search and in the qsearh. But then you double count the leafs of the normal
>>search (same node, same depth in search and qsearch).
>
>Is this the norm for programs? What PostModernist does is to increment abNodes,
>then check if there are any extensions. If (after taking into account any
>extensions) we're out of depth, I *decrement* abNodes and call quiesce(..),
>which increments qNodes as its first step. Total nodes is then abNodes+qNodes,
>and nodes are not counted twice in this scheme.

I do it exactly like you. I think that is sort of the norm now. I think Crafty
does it too. Remains the point: do you count illegal moves, or not? I do count
them. I know some who don't.

>>But: does it make sense to compare NPS between programs? Suppose one program
>>uses SEE pruning and the other program not. Now the program without SEE will
>>have a higher NPS. But does this comparison make any sense?
>>
>
>As you say, it's often comparing apples and oranges. But if we all counted
>nodes in exactly the same way, we'd have taken one variable out of the equation
>when comparing programs.

I agree.

Regards,

Bas Hamstra.

>
>Andrew Williams
>
>
>>
>>
>>Regards,
>>Bas Hamstra.
>>
>>
>>
>>
>>
>>
>>
>>>Uri
>>>
>>>>
>>>>If the answer is negative then we cannot say that one program is a faster
>>>>searcher only because it searches more nodes per second.
>>>>
>>>>We need a clear definition of nodes to compare.
>>>>
>>>>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.