Live blackjack shuffle eats 90s — the true count decays in 12

Live blackjack at most Australian-facing casinos runs on a continuous shuffle machine, and the physical wash takes somewhere between 85 and 95 seconds depending on the table and the dealer. In that window, a true count you built over four shoes is worth almost nothing — the deck you were tracking is gone, and by the time the first hand of the new shoe is dealt, your edge has decayed to roughly zero within 12 minutes of play at 60 hands per hour.

That's not a knock on live dealer blackjack as a product. It's a structural fact about how the format works, and it's the single biggest reason card counters who cut their teeth on RNG blackjack find live tables frustrating. The counting still works in theory. The shoe just doesn't last long enough for the count to matter.

Why 90 seconds is the whole story

A continuous shuffle machine — CSM — doesn't wait for the shoe to run down. Cards go back in as they're played, get randomised, and come out again. In a live dealer studio, the dealer typically does a manual wash and riffle between "shoes" for presentation, and that's where the 90 seconds goes. It looks like a shuffle. It isn't one in any meaningful sense, because the cards were already mixed.

Compare that to a traditional eight-deck shoe at a physical Crown or Star table. You get roughly 416 cards, a cut card placed maybe 75–80% through, and about 60–70 hands of usable penetration. A skilled counter on a 0.5% house edge game with 75% penetration can generate an expected value of maybe 0.8–1.2% of average bet over a shoe. Strip the penetration to zero — which is what a CSM does — and that number collapses.

The 12-minute figure comes from the arithmetic. At 60 hands per hour, 12 minutes is 12 hands. Twelve hands is not enough to rebuild a meaningful true count from a freshly randomised deck, because the running count after 12 hands of an eight-deck game sits somewhere between −3 and +3 in the overwhelming majority of cases, and the true count divides that by roughly 7.7 remaining decks. You're looking at a true count of ±0.4. That's noise.

The penetration problem in plain numbers

Here's the comparison that matters if you're deciding where to play:

The live CSM column isn't zero because counting is impossible. It's zero because the count you build is reset before it can be acted on with any confidence. You need a true count of +2 or better to make a meaningful deviation from basic strategy, and you need it to persist for enough hands to size your bets up. Twelve hands doesn't get you there.

What the 12-minute decay actually looks like

Run the simulation in your head. You sit down at a live table with a $10 minimum and a $500 maximum — typical for an Australian-facing live studio. First 12 hands, the count drifts. Maybe it goes positive, maybe negative. You're flat-betting because you have no signal. Around hand 15, if the count happens to hit +2, you raise to two units. By hand 20, the cards have cycled through the CSM and the count is back near zero.

You've had maybe five hands of mildly positive expectation. At a 0.5% base house edge and a 1% average edge during those five hands on a $20 average bet, your expected profit is about $1. Subtract the $10 minimum you've been bleeding on the other 15 hands at 0.5% edge — about $0.75 — and you're roughly break-even before you account for the fact that you're playing a game with a house edge at all.

The maths doesn't work. It never did. The only reason card counting worked in the first place was penetration, and CSMs removed it.

Where the count does still matter

There's a narrow exception worth knowing. Some live blackjack tables — particularly in studios that run a physical shoe rather than a CSM — do offer real penetration, usually 50–65% on a six or eight-deck game. You can spot these because the dealer will place a cut card and the discard tray will visibly fill. If you find one, the count works, but the window is shorter than a physical casino and the bet spread is usually capped tighter.

The other exception is side bets. Some live tables offer Perfect Pairs or 21+3, and those have their own counting dynamics that don't depend on shoe penetration in the same way. But the house edge on those side bets is typically 2–7%, which is a different conversation entirely.

The dealer's pace is the real variable

One thing that surprises people moving from RNG to live: the hands per hour is not fixed. A fast dealer on a live table can push 80–90 hands per hour. A slow one, especially with chatty players or a studio with a lot of camera cuts, drops to 40–50.

That matters for the 12-minute decay because it's really a hands-based number, not a time-based one. At 80 hands per hour, the decay window is closer to 9 minutes. At 45 hands per hour, it stretches to 16. Neither is long enough to build a meaningful edge on a CSM table, but it does mean the "12 minutes" figure in the title is a midpoint, not a law.

If you're tracking your own results — and you should be, if you're playing seriously — log hands per hour alongside your results. A table running 85 hands per hour at a 0.6% house edge will cost you roughly $51 per hour on a $100 average bet. The same table at 45 hands per hour costs $27. Pace is a bigger variable than most players account for.

What this means if you're playing live blackjack in Australia

The practical takeaway isn't "don't play live blackjack." It's that live blackjack is a basic-strategy game, full stop. The counting edge that exists on physical shoes — and it does still exist, at Crown Melbourne, at Star Sydney, at the smaller venues — does not transfer to the CSM tables that dominate the online live studios.

If you want to count, you need to find a live table with a real shoe and visible penetration. They exist, but they're a minority of the live blackjack inventory, and the bet limits on them are usually tighter than the CSM tables because the operators know exactly what they're offering.

The open question is whether the live dealer format will ever move back toward physical shoes at scale. The CSM is cheaper to run, requires less dealer training, and eliminates the counting edge entirely — which is precisely why it's the default. But there's a segment of players who want the shoe, and the operators who offer it are quietly charging for the privilege through lower table limits.

Worth watching. In the meantime, if you're at a live table and the dealer is washing cards for 90 seconds, you already know what that means for your count. Play basic strategy, keep your bets flat, and treat the whole thing as entertainment with a known cost per hour. That's the honest version of the game.