Movies made it look glamorous. Modern casinos made it brutal. Here's the honest math.
Hollywood loves card counting. Rain Man, 21, The Hangover — genius math nerds cleaning out casinos with their brains. Here's the reality: card counting is real, it does work, and for 99% of the people who learn it, it's not worth the effort. Let's talk about why.
Card counting isn't memorizing every card that's been played. Nobody does that. What counters actually do is track the ratio of high cards to low cards left in the deck using a simple running tally.
High cards (10s and Aces) favor the player — they make blackjacks, help you win double-downs, and cause the dealer to bust. Low cards (2–6) favor the house. When more high cards remain in the deck, your odds improve. Counters just bet more in those situations.
There are dozens of counting systems, but Hi-Lo is the gold standard for beginners and pros alike — simple enough to run in your head under pressure, accurate enough to actually win.
Every card you see gets one of three values:
You watch every card as it's played and keep a running count in your head. That's it. When the count is positive, high cards are still coming. When it's negative, the deck is top-heavy with low cards.
Watch six cards get dealt. See if you can keep up:
Here's where most amateurs mess up. The running count alone isn't enough. A +6 running count means something very different at the start of a 6-deck shoe than it does when only one deck is left to play.
True Count = Running Count ÷ Decks Remaining
If the running count is +6 and there are 3 decks left in the shoe: 6 ÷ 3 = True Count of +2. If only 1 deck is left: 6 ÷ 1 = True Count of +6 — much more valuable.
The higher the true count, the bigger your edge — and the bigger you should bet. Most counters use a schedule like: bet minimum when true count is below +2, bet 2×, 4×, 8× min as the count climbs.
Time for the numbers that nobody wants to hear.
Read those numbers again. One percent. Not ten. Not five. One. And to sustain even that, you need to be nearly perfect — miscounts of just a few cards per shoe eat most of the edge.
The 1970s were the golden age of card counting. The 2020s are not. Casinos have spent fifty years engineering countermeasures, and most of them work.
Cards are reinserted and shuffled after every hand. There is no "deck composition" to track. Card counting is mathematically impossible at these tables. Low-stakes tables increasingly use them.
Dealers now shuffle earlier in the shoe — sometimes after only 50% of cards are played. The deeper you can count, the more accurate your edge; shallow penetration kneecaps you.
Every mid-to-large casino has pit bosses and surveillance specifically trained to spot counters. Big bet jumps, weird bet spreads, players who refuse hit signals — all flagged fast.
You'll be politely "asked to leave" (a backoff) or permanently trespassed. Many casinos share databases through services like Biometrica — get banned at one, you're flagged at dozens.
A single rule change — paying blackjack at 6-to-5 instead of 3-to-2 — triples the house edge and turns counting into a losing proposition. Most low-limit tables are now 6:5.
A skilled counter playing perfectly at decent stakes makes roughly $15–$30 per hour on average, with massive variance. You'll have winning nights of $1,000 and losing nights of $1,500. A good part-time job pays similar money with none of the stress, variance, or risk of being permanently banned from entertainment venues.
For the vast majority of casual players: learn basic strategy instead. It's easier, it's legal everywhere, it can't get you banned, and the hourly expected loss is small enough that blackjack becomes cheap entertainment.
Card counting is a real math advantage — and a real job. If you're not ready for the second part, you shouldn't bother with the first.