Author: Robert Hyatt
Date: 18:10:51 08/13/05
Go up one level in this thread
On August 13, 2005 at 16:46:05, Alvaro Jose Povoa Cardoso wrote:
>On August 13, 2005 at 16:33:31, Robert Hyatt wrote:
>
>>On August 13, 2005 at 16:28:25, Alvaro Jose Povoa Cardoso wrote:
>>
>>>Hi,
>>>Suppose I have and array current_material() wich holds the material score for
>>>every ply and is updated after the call to domove(). Further more, the sign of
>>>it's value is independent of the stm (side to move). If it is positive it means
>>>white's advantage, if it is negative it's black's advantage.
>>>
>>>If after I do a move at the current ply I want to compare the current material
>>>with then material score at the root, how do I do that? Please note I wan't to
>>>do the compare at every ply and not only if the current stm is the same as the
>>>root's stm.
>>>I know I have to fiddle with both the current stm and the root stm, but how?
>>>
>>>should I do?:
>>> at the root:
>>> if (wtm) {
>>> root_material = current_material(1);
>>> }
>>> else {
>>> root_material = -current_material(1);
>>> }
>>>
>>> at the current ply:
>>> if (wtm) {
>>> cur_material = current_material(ply);
>>> }
>>> else {
>>> cur_material = -current_material(ply);
>>> }
>>>Can I now compare 'cur_material' and 'root_material' without worries on their
>>>sign compatibility?
>>>
>>>best regards,
>>>Alvaro Cardoso
>>
>>
>>Since +=good for white in your scheme, always, there is no "sign compatibility"
>>issues. You just have to know that at BTM plies, negative material scores are
>>good, and vice-versa...
>
>But shouldn't I have to take into account wich stm is at the root?
>I just want to take two variables and compare them immediately.
>I do not want to have compare code for white and compare code for black
>
>Alvaro
I don't see why... since the score is relative to white to move... of course
inside the tree, at btm positions you have to realize that - material scores are
good, not bad...
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.