Zied had e-mailed me a .rar archive of his 'games' directory about a year ago (must have been a link to it on some file-sharing host, as it was rather large). Although in the end it proved more convenient to just grab the files directly from the musketeer.net website, and I only used the archive to ...

Oh, apart from Chexx I had all these in my local Jocly git repository already. Did you redo them from scratch? Several of them use modified versions of base-model.js, rather than locating the required patches in their own model file, and redirecting the calls to the standard chessbase functions to t...

Note that numerological approaches to piece-value prediction like this are infamous for their unreliability. The underlying 'theories' contain so many free parameters that it is always very easy to tune it such that you get it exactly right for the 4 orthodox pieces. But then it usually utterly fail...

Why the ridiculously low version of the Hawk (in the rules section)? This must be a typo. I guess most of the pieces listed with value 5 were copy-pasted and not updated to any real value. (E.g. Crowned Rook is definitely way stronger than the Crowned Bishop). But 2.3 for a leaper with 16 moves...!?

I set up an applet where one can practice elementary end-games of King + piece vs King, for many different pieces, including most Musketeer pieces. See http://hgm.nubati.net/rules/EGT.html . The page starts building an end-game table for the selected piece, and then will reply to any move you play o...