Single Comment

Because I am still struggling to implement the Falcon in Joker80, where efficiency is a hallmark, I decided to add a few lines of code to Fairy-Max to implement support for multi-path moves. Fairy-Max is inefficient anyway, and does not know about pins and check tests; it simply plays on until the King is captured.

So it is possible now to define pieces like Falcon in Fairy-Max (in this as yet unreleased version), so that I could already start running some games for asymmetric play testing.

The initial results suggest that a Falcon is not worth nearly as much as mensioned somewhere below. As the Falcon seems a piece similar to the Rook, initially hard to use on a crowded board, but reaching its full potential as the board gets empty, I decided to test it against Rooks. So I took a Capablanca setup, and replaced both Rooks of one side by Falcons. If the Falcon would be really worth 6.5, against a Rook 5, this would mean the Falcon player is leading by 3 Pawns from the outset. Such 'piece odds' games normally produce 80-90% scores. (Simple Pawn odds results in 62% for Capablanca Chess with Fairy-Max.)

The setup seem to be completely balanced, however. Currently it is at 39.5-37.5 for the Falcons, far below the level of significance for determining which piece is better (Rook or Falcon), but almost ruling out completely that the Falcons convey a +3 advantage.

I would currently be inclined to value the Falcon a quarter Pawn above the Rook.