As we discussed in class, I think this is a fairly problematic proposal. It's a bit Pavlovian - you're trying to 'train' the user to respond correctly - further, it's difficult to design an interactive experience that rewards passivity. But let's try to salvage it. Here's one idea.

What if pressing one button (or turning a knob or whatever) brought up a message to do something else with some other button/knob ("Please adjust Parameter X with Knob B" or something). These messages would appear as onscreen instructions superimposed on the backdrop of the videotape image, which would be malfunctioning and getting worse. The user would be therefore be given feedback (keeping him/her engaged) but this feedback would in fact be steering them away from the original goal of actually getting the videotape to play. Establishing that this was in fact the original goal (getting the videotape to play) is the hard part.

And one practical question: how do you get the videotape out of the deck for the next user?

Ehh, too late. I am tossing it... see above post labeled Perry Final #3.