On 11/06/2010 06:33 PM, Alexander Wagner wrote:
Hi,
Back on this topic. Source of trouble is threefold:
1) Scid's analysis did not handle stockfish' (undue?) mate announcements
correctly at all. This has been fixed already in cvs.
2) The analysis filter below is not the one you're looking for. This
filter is there to stop bad move reporting for the player who is in a
lost position (not in a won position!).
Some 65 lines below this one there is another "if the game was won"
filter that is causing trouble. It holds back a comment to a bad move if
it does not spoil the won position. Hence it can be forgiving for not
making this mate-in-one move.
This filter should not be there. Not at all, imho.
3) UCI engines report a mate announcement (instead of a centipawn score)
if they see a mate.
In the old days Scid reported the (arbitrarily high) score of 327 for
any UCI mating line. Actually, internally this is still the case, except
that in the engine window, the 327 score is translated back into a
readable "Mate in 3" text.
For annotation purposes the arbitrary cp score is still used, which
allows simple numerical comparison to understand that a mating line is
better than a non-mating line.
Comparing different mating lines (especially the length of those lines)
cannot be not done effectively using the cp score alone, since they all
have the same score of 327.
Topcs will be addresses in my next annotation update that I am testing now.
For now you may want to test your mating ability by removing the "if the
game was won" filter (or go for the comment-all-moves option) and use a
non-uci engine.
Cheers,
Joost.
> On 10/27/10 19:06, Daniel Karlsson wrote:
>
> Hi!
>
>> Browsing through the source, I found a code snippet which appears to
>> remove annotations for "dead" games. This might be why Scid is not
>> adding annotations in your game.
>>
>> analysis.tcl:
>> # if the game is dead, and the score continues to go down,
>> don't add any comment
>> if { $prevscore> $::informant("++-")&& $tomove == "white" || \
>> $prevscore< [expr 0.0 - $::informant("++-") ]&&
>> $tomove == "black" } {
>> set isBlunder 0
>> }
>>
>> You could try changing Options> Game Information> Configure Informant
>> Values and increase the threshold for The game is considered won "++-".
> In Matthews given example this should of course be the source of
> trouble. I wonder, however if it is sensible, given the fact that Scid
> will not stop the analysis as such, but keep on. All the above does is
> hindering Scid on telling the results.
>
> Plus: in Matthews case he can't work around it easily as the dialogue
> allows only up to 9.9 pawns to be defined for a won game. However, he's
> up 25 or the like already several moves earlier.
>
> Should we therefore remove this condition to help more average players?
> Any opinions?
>
------------------------------------------------------------------------------
Centralized Desktop Delivery: Dell and VMware Reference Architecture
Simplifying enterprise desktop deployment and management using
Dell EqualLogic storage and VMware View: A highly scalable, end-to-end
client virtualization framework. Read more!
http://p.sf.net/sfu/dell-eql-dev2dev
_______________________________________________
Scid-users mailing list
Scid-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/scid-users
