Blackjack Basic Strategy: The Mathematically Correct Way to Play Every Hand

What basic strategy actually is

Blackjack basic strategy is not a system, a betting progression, or a "feel" for the cards. It is a lookup table — a complete prescription for every legal player decision given two inputs: your hand and the dealer's upcard. Follow it exactly and you are playing the game as well as it can be played without tracking which cards have left the shoe.

The table did not come from intuition. In 1956, Roger Baldwin, Wilbert Cantey, Herbert Maisel and James McDermott — four U.S. Army mathematicians using mechanical calculators — published The Optimum Strategy in Blackjack in the Journal of the American Statistical Association. Their work was refined by Edward Thorp in the 1960s and re-derived countless times since with modern computer simulation. The output is remarkably stable: the modern S17 six-deck chart your phone shows you is essentially the 1956 chart with a handful of corrections in the soft-double region.

Every cell of that chart was computed by enumerating the future of the hand — every card that could come, every dealer draw it could trigger, every payout that could result — and choosing the action with the highest expected value. There is no opinion in basic strategy. There is only arithmetic.

How basic strategy reduces the house edge

The house's built-in advantage in blackjack comes almost entirely from one rule: the player acts first. If you bust, you lose immediately, even if the dealer goes on to bust the same hand. That single asymmetry is worth roughly 8% to the house against a player who simply mirrors the dealer's "hit anything 16 or below" routine.

Basic strategy claws nearly all of that back by making decisions the dealer cannot. You stand on stiff hands when the dealer is weak. You double down when your two cards give you a price advantage you will not see again. You split pairs into two hands when each starting card is worth more alone than together. The result, on a standard six-deck game where the dealer stands on soft 17, is a house edge that Wizard of Odds and every other peer-reviewed source put at approximately 0.5%.

The gap between "I sort of know the rules" and "I have memorized the chart" is roughly 1.5 percentage points of edge — far larger than the gap between basic strategy and counting. Memorization is by far the highest-leverage thing a recreational player can do.

The five decisions you can make

Every blackjack hand reduces to a choice among at most five actions:

• Hit (H): take another card.
• Stand (S): keep your current total.
• Double down (D): double your bet, take exactly one more card, and stop.
• Split (P): on a pair, separate the two cards into two hands with a matching bet on each.
• Surrender (SU): fold the hand for half your bet, where the house allows it.

Some decisions are conditional. You can only double on your first two cards (and at some tables, only on totals of 9, 10 or 11). You can only split a pair. Surrender, where offered, is almost always "late" — you check the dealer's hole card for blackjack first, then decide. Basic strategy charts encode all of this; you do not need to remember the table rules separately, only which chart matches the table you are sitting at.

How to read a strategy chart

Every basic strategy chart is laid out the same way: your hand down the left side as rows, the dealer's upcard across the top as columns (2 through Ace). Find the intersection, take the action.

The charts are usually split into three blocks because the math differs:

• Hard totals — hands without an Ace, or with an Ace that must count as 1 to avoid busting.
• Soft totals — hands containing an Ace counted as 11.
• Pairs — two cards of the same rank, where splitting is on the menu.

If you only ever look at one chart, look at the hard totals chart. It covers the majority of hands you will be dealt and contains the decisions where mistakes cost the most.

Hard totals: the spine of the chart

Hard totals are where intuition fails most often, because the correct play depends almost entirely on what the dealer is showing — not on what you have.

Consider three textbook examples:

• Hard 16 vs dealer 10. The worst hand in blackjack. You will lose this hand about 77% of the time no matter what. Basic strategy says hit (or surrender, if available) because standing is even worse — you win only when the dealer busts, and a 10 upcard busts roughly 23% of the time. Hitting at least gives you a small chance of improving.
• Hard 11 vs Ace. Counterintuitive to many players: against the dealer's strongest card, you still double. Eleven is a price you will rarely see again, and a 10-valued card is the most common next card in a 52-card deck. In an S17 game you double; in an H17 game it becomes even more attractive.
• Hard 12 vs dealer 2 or 3. You hit. The dealer's 2 and 3 are usually called "bust cards", but they actually bust less than 40% of the time. Standing on 12 wins only on a bust, so you take the hit and accept the ~31% chance of busting yourself.

The pattern most people internalize: stand on 12-16 when the dealer shows 4, 5 or 6; hit them when the dealer shows 7 through Ace. That single rule, applied mechanically, gets a player most of the way to correct play on hard totals.

Soft totals: the part everyone gets wrong

A soft hand contains an Ace counted as 11. The Ace's flexibility is the entire point — you cannot bust on the next card, so you can afford to be aggressive in ways a hard hand would not allow.

Two soft-hand decisions trip up nearly every untrained player:

• Soft 18 (A,7) vs dealer 9, 10 or Ace. Most players stand here. They have 18 — a winning hand most of the time, right? Wrong. Against a 9, 10 or Ace the dealer will end up with 19 or better far more often than not, and your 18 loses. Basic strategy says hit. Standing on A,7 vs 9 is one of the most expensive common mistakes in the game.
• Soft 17 (A,6) vs dealer 6. Double. You are not improving to a great hand most of the time, but the dealer has the worst possible upcard and the double extracts maximum value from their weakness. A,6 should almost never be played as a stand-only 17 — it is a hitting or doubling hand, never a final total.

Pair splitting: when two hands are better than one

Two rules in pair splitting are absolute:

• Always split Aces. Hard 12 plays terribly. Two separate hands each starting with an 11 play wonderfully. The math is so lopsided that casinos restrict the play — you usually get only one card on each Ace, and a ten on a split Ace counts as 21, not a blackjack.
• Always split 8s. A pair of 8s is a hard 16, the worst hand in blackjack. Splitting turns one losing hand into two starting hands of 8 — not great, but mathematically better against every dealer upcard, including a 10 and an Ace. You are not splitting 8s to win; you are splitting them to lose less.

Most other pairs are situational: split 2s, 3s, 6s, 7s and 9s only against specific dealer upcards; never split 5s (you have a 10, double or hit it); never split 10s (you have 20, the second-best hand in the game). For the complete logic, see when to split in blackjack.

Surrender: cut your losses correctly

Surrender — when offered — is one of the most undervalued plays in casual blackjack. By forfeiting half your bet, you escape hands whose expected value is worse than −0.5 units. The list of late-surrender hands is short and worth memorizing as a block:

• Hard 16 vs dealer 9, 10 or Ace.
• Hard 15 vs dealer 10 (and vs Ace in H17 games).

That is it for standard six-deck rules. If your table does not offer surrender, ignore the option entirely. If it does and you skip it, you are leaving roughly 0.07% of edge on the table — small per hand, large over a lifetime. Full breakdown at when to surrender.

Rule variations change the chart

There is no single basic strategy chart — there are several, each tuned to a specific rule set. The differences are small but real:

• S17 vs H17. When the dealer hits soft 17, you double a few more borderline hands (A,8 vs 6, 11 vs Ace) and surrender a few more 15s and 17s against an Ace. The edge shift is about +0.22%. Details at S17 vs H17.
• DAS vs no DAS. If you cannot double after splitting, several borderline pair-splits stop being profitable.
• Number of decks. Single-deck strategy diverges from six-deck in roughly a dozen cells. Eight-deck and six-deck are nearly identical.
• European no-hole-card (ENHC). The dealer takes a hole card only after the player acts. You play more conservatively against 10s and Aces because the dealer cannot have already busted to a blackjack.

For a typical player, the answer is: learn one chart (six-deck S17 DAS, late surrender) cold, then adjust two or three cells when you sit at an H17 table. Do not try to learn five charts at once.

Why you must actually memorize it

You cannot pull up a strategy app at the table. Most casinos allow printed strategy cards but frown on phones, and even at a friendly home game the pace of the game punishes slow decisions — dealers, pit bosses and other players read hesitation as inexperience, which invites scrutiny and breaks your own concentration.

More importantly, the cost of a single wrong decision is real money. A player who plays "near-perfect" basic strategy at, say, 95% accuracy is giving up roughly 0.2% of edge on top of the baseline 0.5% — doubling the house's take. Memorization is not perfectionism. It is the difference between a game with a half-percent edge and a game with one percent.

Spaced repetition, drilling weak spots, and structured chart review are how serious players close that gap. See how to memorize basic strategy for a complete study plan.