When a blindfold expert plays several players simultaneously, there is a tacit assumption that the players will not collude. But suppose that the players are allowed to collude, and their goal is to maximize the chances of making the expert lose track of or forget the exact configuration on at least one of the boards. (I'm assuming that they do this not by indulging in unsportsmanlike behavior unrelated to chess itself, but only by judicious move selection.) What strategy should they adopt?
I suspect that a good strategy would be to choose very similar but slightly different moves, like playing b6 and g6 and then shuffling the bishops back and forth in some random manner. If the expert is using memory palaces or some such technique then it will be challenging to come up with mnemonics that accurately distinguish between such similar positions.
Have any blindfold experts prepared for the possibility of such collusion? What strategies worry them most? How many simultaneous blindfold games would be possible if this kind of collusion were allowed?