Winboard Forum

SMIRF had calculated the over all position evaluation using the well known average piece exchange model values. During the last time I experimented with delegating the mobility parts of those values to the detail evaluation part. Thus there have been "-X" versions for a mixing of those approaches and "-0" for completely avoiding the mobility parts.

In short there will be a new "MS-174b-0" bonus version of SMIRF, which should play a little bit stronger then - as I hope ...

Currently I have integrated a new component of an evaluation function in SMIRF. It is the Elephantiasis-Reduction as currently discussed in the http://www.chessvariants.org/index/listcomments.php?subjectid=Piece+Values forum, concerning piece values. Still being in my private version, it will be released soon within the bonus version of SMIRF. This new element is useful in battles between different armies. I also noticed a speed up in mating tasks of traditional 8x8 and 10x8 CRC positions.