Futzing around with some cards recently, I was reminded of an observation I first made some 20 years ago: namely, that a move need not be "finger-perfect" to be deceptive and that occasionally, a "finger-imperfect" move is superior to finger-perfect alternatives.
I originally observed this while studying Slydini's version of the Equal/Unequal Ropes (which I REFUSE to call Professor's Nightmare). Slydini devised a brilliant false count, in which the ropes are pulled horizontally, one-by-one from the left hand into the right. What makes it especially deceptive is how the already-counted "first" rope hangs vertically and lifeless as the second rope is pulled out of the left hand. What fascinates me is that the relationship between objects and fingers during the count is in fact completely illogical. On the count of two, the "first" rope is seen to be held by different fingers than were holding it on the count of one. Yet the visual narrative of the count, and the smoothness with which it can (should!) be executed, trumps the illogical fingering, even to rapt close-up observers.
I was reminded of this several months ago, while devising a stand-up version of "Le Temps Four Aces" from Hugard/Braue. After much experimentation, I came up with (though wouldn't dare claim to have originated) a top-change in which three spider-gripped aces are substituted for three indifferents. In fact, the spatial relation of cards to fingers is markedly different after the switch than before. Yet the smoothness of the move, and the especially "gingerly" way the cards are held after the switch, make the move more convincing than the more "finger-perfect" variants I toyed with.
I'm sure I'm not the first to make this finger-perfect/finger-imperfect observation. Given the limited nature of my magic library, maybe others have written about it at length. I just think it's a cool, subtle point of magic theory and wonder if others have thoughts on the matter.