Most casino games give a player no meaningful decisions at all - a slot spins, a roulette ball lands, and no choice made after the bet changes the outcome. Blackjack is different. Every hand presents a genuine decision - hit, stand, double, split, or surrender - and the difference between the mathematically correct play and an instinctive one adds up to a measurable, real difference in long-run results. This guide covers exactly how to make that correct decision for every common hand, why it works, and how table rules shift the numbers underneath it.
What "basic strategy" actually means
Basic strategy is the mathematically optimal play for every possible player hand against every possible dealer up card, calculated by weighing the probability of every outcome for both hitting and standing (and, where relevant, doubling, splitting, or surrendering) across a large number of simulated hands. It was first derived by hand in the 1950s and later refined using computer simulation, and it remains essentially unchanged today for a given set of table rules - the "correct" play for a hard 16 against a dealer 10 doesn't shift with trends, because it's a fixed mathematical answer rather than an opinion.
Following basic strategy consistently reduces the house edge in blackjack to somewhere around 0.5%, among the lowest of any widely available casino game - far below the roughly 2.7-5.26% carried by roulette, for instance, covered in Roulette Strategy. Playing on instinct instead of basic strategy typically pushes the effective house edge up to somewhere between 2% and 5%, depending on how far off the instinctive plays are from the correct ones - which is to say, basic strategy alone is worth more to your actual results than almost anything else you can do at a blackjack table.
Hard hands
A "hard" hand is any hand without an Ace, or with an Ace that can only count as 1 without busting. Hard hand strategy is where the biggest, most consequential decisions happen, and a few rules cover the large majority of situations:
- Hard 8 or less: always hit - the hand is too weak to stand on regardless of the dealer's up card
- Hard 9: double against dealer 3-6, otherwise hit
- Hard 10: double against dealer 2-9, hit against dealer 10 or Ace
- Hard 11: double against every dealer up card in most rule sets (some conservative charts call for hitting against an Ace specifically)
- Hard 12: stand against dealer 4-6, otherwise hit
- Hard 13-16: stand against dealer 2-6, otherwise hit
- Hard 17 or more: always stand
The logic underneath these rules is consistent: dealer up cards of 4, 5, and 6 are the dealer's weakest position, since house rules force the dealer to keep drawing on a stiff total and those up cards bust at a meaningfully higher rate. Standing on a marginal hard total (12-16) against a weak dealer up card lets the dealer's own bust risk do the work; hitting the same total against a strong dealer up card (7 through Ace) is necessary because standing on a weak hand against a strong dealer hand loses far too often to be worth avoiding the bust risk.
Soft hands
A "soft" hand contains an Ace counted as 11 without busting - soft 18, for instance, is an Ace and a 7. Because the Ace's flexibility means a soft hand can never bust on the next card (worst case, it simply converts to its hard equivalent), soft hand strategy is meaningfully more aggressive than the equivalent hard total:
- Soft 13-14 (A,2 / A,3): double against dealer 5-6, otherwise hit
- Soft 15-16 (A,4 / A,5): double against dealer 4-6, otherwise hit
- Soft 17 (A,6): double against dealer 3-6, otherwise hit - standing here is a common beginner mistake
- Soft 18 (A,7): double against dealer 3-6, stand against 2, 7, or 8, hit against 9, 10, or Ace
- Soft 19 and higher: stand
Soft 17 and soft 18 are the two most commonly misplayed hands in blackjack, precisely because "17" and "18" feel safe by hard-hand instinct. Since a soft hand can't bust on the next card, that instinct is actively working against you on these two totals specifically.
Pair splitting
When your first two cards are a pair, you can split them into two separate hands, each receiving its own additional card and requiring a separate equal wager. A few pairs have a single, unambiguous correct answer regardless of the dealer's up card:
- Always split Aces and 8s. A pair of Aces is one of the strongest starting positions in the game, and splitting maximizes the chance of forming two strong hands rather than one mediocre 12 (or worse, a hand awkwardly forced to treat one Ace as 1). A pair of 8s totals a weak hard 16 - one of the worst hands in blackjack - and splitting converts that single bad hand into two hands each starting from a much stronger 8.
- Never split 10s or 5s. A pair of 10-value cards is already a strong 20; splitting it turns one strong hand into two uncertain ones, giving up a near-guaranteed win for a coin flip. A pair of 5s totals a hard 10, one of the best doubling hands in the game - splitting it into two hands starting from 5 each throws away a strong doubling opportunity for a much weaker starting position.
- Split 2s, 3s, and 7s against dealer 2-7, otherwise hit.
- Split 6s against dealer 2-6, otherwise hit.
- Split 9s against dealer 2-6 and 8-9, stand against dealer 7, 10, or Ace (since a pair of 9s already totals a strong 18 against a dealer 7, which is likely already a winning hand).
- Never split 4s in most standard rule sets - a hard 8 is a reasonable hand to hit or double from, and two hands starting from 4 each is generally worse.
Whether doubling after a split (DAS) is allowed affects several of these calls at the margins - always check the specific table's rules before assuming a standard chart applies exactly.
Doubling down
Doubling down lets you double your original wager in exchange for exactly one additional card, with no further hits allowed on that hand. The strongest doubling opportunities cluster around a dealer showing 4, 5, or 6 - the dealer's weakest position - combined with a player hand strong enough that one more card is likely to produce a genuinely strong total. Hard 10 and hard 11 are the two best hard-hand doubling opportunities in the entire game; soft hands from soft 13 through soft 18 all carry meaningful doubling windows specifically because they carry zero bust risk on the extra card.
Surrender
Where offered, surrender lets you forfeit half your wager immediately, before playing the hand out, in situations where the odds are heavily stacked against you regardless of how you play. Not every table offers surrender, and where it's offered, it's typically "late surrender" (available only after the dealer checks for a natural blackjack). The clearest correct surrender spots are a hard 16 against a dealer 9, 10, or Ace, and a hard 15 against a dealer 10 - situations where the expected loss from playing the hand out exceeds the guaranteed 50% loss from surrendering. Surrender is a small, situational tool rather than a frequently used one, but skipping it entirely when it's available and correctly indicated leaves real value on the table.
Memorizing a strategy chart efficiently
Trying to memorize an entire basic strategy chart in one sitting rarely sticks. A more effective approach is learning it in blocks: master hard hand totals first, since they come up most frequently and cover the largest share of decisions you'll actually face; then soft hands, which diverge sharply enough from hard-hand intuition that they benefit from being drilled separately rather than blended in; then pairs, which have the fewest genuinely ambiguous decisions once you know the "always split" and "never split" rules above. Practicing with a free demo-mode table - covered in Casino Games Explained - lets you apply the chart against real hands without financial risk while it's still becoming automatic, which is a far more effective way to internalize it than memorizing rows and columns in isolation.
Why table rules matter as much as your own play
Two blackjack tables can look identical at a glance and carry meaningfully different house edges purely due to their rules. The variables that matter most:
| Rule | Favorable version | Impact |
|---|---|---|
| Blackjack payout | 3:2 | 6:5 tables raise the house edge by roughly 1.4%, among the single worst rule changes possible |
| Dealer soft 17 | Dealer stands | Dealer hitting soft 17 adds roughly 0.2% to the house edge |
| Number of decks | Fewer decks | Single- and double-deck games carry a slightly lower house edge than 6-8 deck shoes, all else equal |
| Doubling after split (DAS) | Allowed | Disallowing DAS costs the player roughly 0.1-0.15% |
| Surrender | Offered | Late surrender availability shaves a small additional edge for the player when used correctly |
Of everything in this table, the blackjack payout ratio matters most and is worth checking before sitting down at any table - a 6:5 payout on blackjack (rather than the standard 3:2) is a meaningfully worse deal even at an otherwise identical table, and it's worth actively avoiding in favor of a 3:2 table at a similar limit wherever one is available.
Card counting: why it doesn't apply to most online play
Card counting - tracking the ratio of high to low cards remaining in the shoe to gain a genuine, if modest, edge over the house - is a well-documented, legitimate strategy in physical casinos with a shoe that depletes over a session. It doesn't apply to standard RNG-based online blackjack, since the virtual deck is reshuffled before every hand, meaning there's no depleting composition to track from hand to hand. Live dealer blackjack uses a real, physical shoe in principle, but studios typically use a large number of decks and shuffle relatively frequently, making traditional counting far less practical than it would be at a single-deck physical table - see our Live Dealer Guide for more on how the format works. For the large majority of readers playing online, basic strategy - not counting - is where the real, accessible edge lives.
Blackjack side bets
Many blackjack tables - live and RNG alike - offer optional side bets alongside the main hand: Perfect Pairs (betting your first two cards form a pair), 21+3 (combining your first two cards with the dealer's up card into a three-card poker hand), and others. These are entirely separate wagers from the main blackjack hand, resolved independently of whether you win or lose the hand itself, and they typically carry a meaningfully higher house edge than a well-played main hand - often in the 3-8% range or higher, well above blackjack's roughly 0.5% edge with correct basic strategy. Side bets aren't inherently a mistake to play, but they're worth understanding as a distinct, higher-variance addition rather than a core part of optimizing your blackjack results, and they shouldn't be sized using the same bankroll logic you'd apply to your main hand wager - see Bankroll Management for how to think about sizing bets with meaningfully different house edges within the same session.
Common mistakes
Standing on soft 17 or soft 18 out of habit. Since a soft hand can't bust on the next card, the "17 is safe" instinct from hard-hand play actively works against you here.
Taking insurance. Insurance - a side bet offered when the dealer shows an Ace, paying 2:1 if the dealer has blackjack - is a net-negative bet for a player not counting cards, regardless of your own hand's strength, since the base probability of a dealer blackjack doesn't justify the payout offered.
Playing 6:5 tables without checking the payout first. This is the single most costly rule variation a player can unknowingly sit down at, and it's worth a five-second check before your first hand at any new table.
Splitting 10s because "20 feels safe to risk." A made hard 20 is already a strong, likely-winning hand; splitting it into two independent hands is a statistically worse decision in almost every situation, regardless of how tempting a pair of aces-adjacent-value cards might feel.
Where to go from here
Once basic strategy feels comfortable, the next highest-leverage thing to get right is how much to actually wager on each hand - covered in full in Bankroll Management. For the broader picture of how blackjack compares to other casino games on house edge and skill involvement, see Casino Games Explained, and for how we verify a specific blackjack table's actual rules and payout structure before publishing a review, see How We Review Casino Games.