This I was contemplating to make as my next research project after I finish the one I've been working on for a few months.
Since shuffles are far from random, the consistency of a previous shoe may (or may not) in theory be able to predict the EV of the next shoe without any shuffle-tracking.
One would think (and I'd agree) that it would be preferable for the shoe to be as slugged as possible (within practical limits). If a slug of high cards is located near the beginning of a shoe, then you'll see the good cards, and be able to promptly wong-out for a phone call or something of the like. If a slug of high cards is found near the end of the shoe (but not behind the cut-curd), then this is also desirable, since it is something you'd likely have been able to predict through counting. It follows that if your count had predicted the high-cards to be coming then you'd have big bets out during this segment.
To the contrary, if the shoe never deviated to a TC greater in magnitude than +/- 1, a de facto CSM shoe if you will, then there is no way we could win by counting alone, assuming we are not wonging. If you never see significant advantage, you can't place large bets to make up for the small ones we've expected to lose.
So, the question becomes this: How much does the slug-character of the previous shoe influence the next shoe? Remember that as slug-character from the previous shoe can be diluted, but new ones can form or slugs may be combined. (It, however, by law or entropy will move towards randomness on average, but maybe not much if a weak shuffle is used).
The joint I play at, I believe uses a 1 pass R&R with plugs (It actually could be a multi-pass, but I didn't start learning about ST'ing until after my last visit there, which was about a month ago, and thus did not pay much attention to the shuffle). Regardless, though, I just highly doubt that 1 pass R&R does too much to affect the slug-character of the following shoe.
So, the question becomes, for non-ST'ers, could it be possible, if you noticed that the TC fluctuation was minimal during the previous shoe to predict that the EV of the next shoe will be significantly lower than the EV of the average shoe? This information could possibly be useful to perhaps wong out before the shoe even begins if you know it is expected to be uneventful.
SP