I now ran some matches with the individual Zen-Zebra pieces against their FIDE counter-parts, in a FIDE context. Only the Charging Zebra was able to beat the Bishop pair, with a moderate 54.5% score over 199 games (where the statistical error should be 2.8%). Such a score corresponds to a quarter of a Pawn, and that for a pair of pieces.
The other Zen pieces lost, sometimes badly. The FIDE Rooks won with a 66% score, about the same as Pawn odds, suggesting they are each 0.5 Pawn better than their replacements. The Knights won by 56.5% over 100 games (4% error), which translates to a 0.15 Pawn superiority of a single Knight. Perhaps somewhat surprisingly the Queen replacement did very bad: the FIDE Queen won with a 59% score, slightly over half a Pawn, and that for only a single replacement.
This is all very tentative, as I only run the tests with white suffering the replacements (which normally should have given them the advantage), and I did run all games from the same initial position, and did not check for duplicat games. But is seems especially the Rook and Queen replacements fall short of their target.
So it seems the Zebra moves, despite their large leap, are not dangerous enough to compensate for the fact that they more often fall off board. It could also be that the asymmetry of the Z moves in the Zen pieces disadvantages them. It enables them to raid deeply into enemy territory, but they then cannot get out the way they came, and might be trapped there. This desrves an investigation, e.g. by testing FffsbZ (asymmetric) against FvZ and FsZ (symmetric).
I now ran some matches with the individual Zen-Zebra pieces against their FIDE counter-parts, in a FIDE context. Only the Charging Zebra was able to beat the Bishop pair, with a moderate 54.5% score over 199 games (where the statistical error should be 2.8%). Such a score corresponds to a quarter of a Pawn, and that for a pair of pieces.
The other Zen pieces lost, sometimes badly. The FIDE Rooks won with a 66% score, about the same as Pawn odds, suggesting they are each 0.5 Pawn better than their replacements. The Knights won by 56.5% over 100 games (4% error), which translates to a 0.15 Pawn superiority of a single Knight. Perhaps somewhat surprisingly the Queen replacement did very bad: the FIDE Queen won with a 59% score, slightly over half a Pawn, and that for only a single replacement.
This is all very tentative, as I only run the tests with white suffering the replacements (which normally should have given them the advantage), and I did run all games from the same initial position, and did not check for duplicat games. But is seems especially the Rook and Queen replacements fall short of their target.
So it seems the Zebra moves, despite their large leap, are not dangerous enough to compensate for the fact that they more often fall off board. It could also be that the asymmetry of the Z moves in the Zen pieces disadvantages them. It enables them to raid deeply into enemy territory, but they then cannot get out the way they came, and might be trapped there. This desrves an investigation, e.g. by testing FffsbZ (asymmetric) against FvZ and FsZ (symmetric).