I have a chess variant-related question in mind, that I'm not sure if it's on topic because it has too many potential answers and it's not really a practical question about puzzle making (but arguably can be useful for chess puzzle making), more of a challenge in itself. I'm looking for the "best" configuration(s) of fairy chess pieces under certain rules, in the sense that there's a winning strategy against the other configurations or something like that. So an answer can be objectively better than many others, but not all pairs of answers are comparable (no winning strategy known either way), and there might not a single best answer, but I am interested in getting answers that are better than most others even if they can be beaten by specific other answers. It's not subjective because an answer can be objectively better than another, but it's not only about finding the "best" answer, and it's also not an optimization because there's no numeric score to optimize for. (I considered adding other criteria eg. win rate under randomized starting configurations, to determine which answer is better for hard-to-compare pairs, but it requires a lot more computation and may be unfun and impractical to do, and it can't rule out rock-paper-scissors scenarios anyway.)
The question is the following if it matters:
What's the most powerful fairy chess piece possible that is still playable, in the sense that it doesn't necessarily make the opponent start out in an illegal position (eg in check) or have a simple winning strategy no matter how powerful the opponent's army is? Under certain assumptions to avoid trivial cases of course, but I'm still figuring out what the rules should be.
Edit: I saw meta posts like this About NON-open-ended puzzles with multiple answers working towards an optimum and The end of open-ended puzzles about how this site doesn't allow open-ended questions with no single best answer (unless they are about puzzle making or something) and that optimization questions are allowed if they seem to have a provably optimal answer (but I'm new here and am not sure what's the current consensus or how these policies are implemented now). One answer said there are two types of "finding the best X" questions: one where finding an X is a good challenge in itself and one where the challenge is only in beating other answers. Now my question may be rephrased as about fairy chess puzzle making, but it feels like a bit of a stretch. It's not only challenging because of the need of beating others, because first one has to satisfy the conditions and constraints that already rule out many trivial attempts, and then one can reasonably anticipate other possible answers that may "beat" a seeming potential answer (in fact I have done so several times, that's why I think this question is worth asking). Finding an answer that holds up against other obvious candidates is non trivial. Many obvious possible "best" answers are not unbeatable, for instance:
Actress (Q+N+Camel) is probably the most powerful among the better-known existing fairy chess pieces, but it's by no means the strongest, as it can trivially be beaten by universal leaper (leaps to anywhere). But is universal leaper too strong to be playable? Even seemingly omnipotent pieces can be beaten under certain assumptions: universal leaper is not unbeatable if it's the only piece White has and Black has multiple adjacent Siamese Kings (that all need to be captured), or even just a single King that can resurrect itself on an adjacent square. That's why I need to be careful about the ground rules on what abilities and configurations are allowed. I currently think Siamese Kings are allowed in the defender's configuration, as they are a thing in some chess puzzles, so universal leaper is not "unplayable-strong" and is likely not the "best" answer.
