“The card player won’t know by casual play, but he could by investigation.”

“Design by intervention and design by deck-stacking are empirically equivalent after-the-fact.”

The argument rests on this premise, but I wonder if it is true. Would there be no empirical difference whatsoever between intervention and deck-stacking, or only limitations based on current observational abilities? Consider the analogy again: I may not know how my friend did it if I made no examination of the decks, either before or after the round. But say I stopped and counted cards, or checked his sleeves, or something along those lines? Then the empirical evidence may be stronger towards one method of cheating versus another.

Carried, then, from the analogy to the concrete, both methods of design will leave some evidence distinctive to each. One might argue that intervention will leave certain marks on the record of nature that deck-stacking wouldn’t, and vice versa. It’s just a matter of being able to detect them. The card player won’t know by casual play, but he could be investigation. Likewise, the scientist may know by a careful investigation of the evidence, the accuracy of which depends greatly on the tools available.

Maybe another way of saying it is: the empirical equivalence of design methods is in the epistemlological realm (and possibly surmountable) but perhaps not in the ontological realm. What do you think?