First, the game itself — 60 seconds. In blackjack you get two cards and try to beat the dealer's hand without going over 21. Number cards are worth their number, face cards (J, Q, K) are worth 10, and an ace counts as 1 or 11 — whichever helps you. Each turn you can hit (take another card), stand (stop), double down (double your bet for one more card), or split a matching pair into two hands. The dealer has zero choices: they must keep hitting until they reach 17 or more, and if they pass 21 they bust and every player still standing gets paid.

One hand is special: an ace plus any ten-value card as your first two cards is a blackjack, and it pays 3:2 — bet $10, win $15 instead of the usual $10. Remember that payout; a lot of your future edge lives inside it.

In a 6-deck game, the dealer deals from a shoe — a plastic box on the table holding six decks shuffled together, 312 cards. Played with perfect decisions, blackjack's house edge (the casino's built-in mathematical advantage) is only about half a percent — the smallest of any table game. And here's the crack in the wall: blackjack is the only casino game with a memory. Roulette forgets every spin. Slots forget every pull. But every card that leaves the shoe is gone until the next shuffle — what's left inside keeps changing, and what's left decides who has the edge.

Why do leftover tens and aces help you and not the dealer? One: more blackjacks get dealt — yours pay 3:2, the dealer's only takes your bet. Unequal payout, your favor. Two: the dealer is forced to hit hands like 14 and 15, and a ten-heavy shoe busts those hands relentlessly — while you're allowed to just stand and watch. Three: your double downs want a big card, and a ten-rich shoe delivers them.

You will not be memorizing cards. You'll track one number, bet small when the shoe is cold, and bet big when it's hot. The 6-deck shoe is the game every casino in the world deals — master it and you're never without a table.

There are dozens of counting systems with intimidating names. Ignore them all except Hi-Lo — the system used by the MIT teams and virtually every professional who ever mattered. It's the best trade of power for simplicity ever devised, and it assigns every card one of three values:

The logic clicks once you see it: when a 5 comes out, a card the dealer needed just left the shoe, so the count goes UP — good for you. When a king comes out, one of your cards left, so it goes DOWN. Suits mean nothing. A ten and an ace count exactly the same (−1).

The count starts at 0 on every fresh shoe. As cards hit the felt, add their values in your head. That total is your running count. In a 6-deck game it drifts far — you'll see +12s and −10s across a shoe — and it means one thing only: high positive = the shoe is loaded with tens and aces.

One built-in gift: every deck contains exactly twenty +1 cards and twenty −1 cards, so a complete deck counted correctly always ends at exactly 0. Any time you practice with a real deck, the deck itself grades you.

Knowing the values is step one. Keeping up with a real dealer is step two — casino dealers are fast, and a beginner adding card-by-card falls behind by the third hand. The fix is how every pro actually counts: in pairs, because pairs cancel.

A king and a five land together? −1 and +1 = zero — you don't even think a number, you just skip them. A queen and a jack? −2, one single thought. A 4 and a 3? +2, one thought. Roughly a third of all pairs cancel to zero, which cuts your mental work by a third and turns the rest into a rhythm instead of arithmetic.

What you'll actually count: every card you can see, exactly once. Good news for shoe players — in 6-deck games, all cards are dealt face up, right there in front of you (players aren't even allowed to touch them). A full table has seven betting spots, so one round can put fifteen-plus cards on the felt: your hand, six other hands, the dealer's upcard, then every hit card and finally the dealer's hidden hole card when it flips. Develop one consistent sweep — left to right across the table, dealer last — and never count the same card twice.

Here's a problem the running count can't solve alone. Say your count is +6. If those six extra small-cards-gone are spread across five remaining decks — 260 cards — the shoe has barely changed. But the same +6 concentrated in the last two decks? That's a genuinely hot shoe. Identical running count, completely different realities. So counters convert to a per-deck number:

Where does "decks remaining" come from? The discard tray — the clear plastic holder beside the dealer where all dealt cards pile up. You'll estimate by height: a deck of 52 cards stands about 3/4 of an inch tall, two decks about an inch and a half. Six decks in play, tray showing two decks used → four remain. Whole-deck accuracy is fine early in the shoe; tighten to half-decks late, when the count matters most. Nobody's born reading card stacks — it's a one-week skill.

One more term you must know before choosing a table: penetration — how deep into the shoe the dealer deals before the shuffle. Dealers place a colored cut card in the shoe; when it appears, they reshuffle everything, count resets to zero. Deep penetration (cut card near the back) means you spend more rounds with a small divisor and real information — that's where every dollar of counting profit lives. 75%+ penetration = playable. A dealer who cuts off 2 of 6 decks? Marginal. Cut at 50%? Walk — that table cannot be beaten by anyone, no matter how well they count.

Time for the part most beginners get backwards. Counting doesn't make money. Betting more when the shoe favors you and betting the minimum when it doesn't makes money. You'll still lose plenty of high-count hands — the edge is a few percent, not a crystal ball. The profit comes from having your big bets on the felt during the favorable hands, thousands of times over.

First, a word you need: a unit is your base bet — usually the table minimum. At a $10 table, one unit = $10, so "12 units" means $120. In a 6-deck game your edge typically arrives around true +1.5 to +2, and because most of every shoe sits in neutral territory, shoe players need a wide ladder:

Honest expectations: this spread in a decent 6-deck game earns roughly $10–$30 an hour at $10 units. Not movie money — compounding money. And a wide spread at a busy shoe table blends right in, because shoes are full of gamblers pressing streaks. Ride that cover.

Everything so far assumed you sit down and play the whole shoe — grinding minimum bets through the cold stretches, waiting for heat. Now the shoe game's superpower, named after gambling author Stanford Wong: don't play the bad counts at all.

It's called back-counting, and it looks like nothing: you stand behind a table like any casino tourist watching the action — totally normal behavior on a busy floor — while silently counting every card from the fresh shoe. The table is doing your prospecting for you, for free. When the true count climbs to +2, you slide into an open seat, chips out, and start betting real money into a shoe you already know is hot. If the count later sinks below +1, your phone "buzzes" — so sorry, gotta take this — and you walk, straight to the next table's fresh shoe.

The casino's counter-move is a small sign reading "No Mid-Shoe Entry" — meaning new players may only join when a fresh shoe starts, which exists specifically to stop what you just learned. Tables without the sign are your hunting ground. One courtesy that doubles as cover: enter smoothly between rounds, chat with the dealer, look like a guy who just finished his drink and felt lucky.

Two definitions before this lesson makes sense. Basic strategy is the chart of the mathematically best play for every possible hand against every dealer upcard — when to hit 16, when to double 11, when to split 8s. It was computed decades ago, it's free to print online, and memorizing it 100% cold is what gets the house edge down to ~0.5% before counting adds anything. Insurance is a side bet the dealer offers whenever their upcard is an ace — you're betting half your wager that they have a ten underneath for blackjack. For a normal player it's a famous sucker bet. Hold that thought.

Here's the thing: basic strategy assumes a neutral shoe. When your true count says the shoe isn't neutral, a handful of plays flip to the opposite answer. These are called deviations, written like "16 vs 10" — meaning your hand totals 16, the dealer's upcard is a ten. Memorize these ten and skip the rest — they carry most of the value:

In a shoe you'll use deviations less often than you'd think (neutral counts dominate), but when a late-shoe count spikes, deviations + max bets + deep penetration stack together — the hands that make your month live in those final two decks.

The truth no one selling a "system" tells you: with a roughly 1% edge, losing nights are common, losing weeks happen, and losing months are possible — while doing everything right. A $10–$120 spread can drop several hundred dollars in one ugly hour. Variance is violent, and it doesn't care that you counted perfectly. The count guarantees nothing in a session — only across hundreds of hours.

Your bankroll is money set aside purely for play — not rent money wearing a disguise. It's the armor that lets you survive the swings long enough for the edge to show up:

Now meet your opposition. The pit boss is the supervisor standing behind the dealers watching several tables at once; above everyone is the eye in the sky — ceiling cameras with staff trained to spot players whose bets track the count. Counting with your brain is legal — no US law forbids playing well — but casinos are private businesses, and their move is the back-off: a polite "you're welcome to play anything but blackjack," or an invitation to leave. No arrest, no confiscated chips, no scene. Your defense is never being worth a second glance: climb your bets like a gambler pressing a streak instead of jumping 1-to-12 in one move, keep sessions around an hour, talk, tip small, cheer for the table, and never say "count" out loud.

Last habit, non-negotiable: log every session — date, casino, hours, penetration, spread, result. Hundreds of hours from now, that log is the only honest answer to "is this actually working?" — and on the brutal nights, it's what reminds you the math is bigger than one session.