Doubling down is one of the highest-value plays in blackjack — done correctly, it lowers house edge; done incorrectly, it hands the casino free money. The rules are precise: always double 11 against dealer 2 through 10, double 10 against dealer 2 through 9, double 9 against dealer 3 through 6, plus specific soft double situations. Everything else, don't double. This guide covers the math behind each decision and the mistakes that cost real money.
Double on 11: against dealer 2-10 (double vs A if H17, hit if S17 no surrender)
Double on 10: against dealer 2-9
Double on 9: against dealer 3-6
Soft 18 (A,7): double against dealer 3-6
Soft 17 (A,6): double against dealer 3-6
Soft 16 (A,5): double against dealer 4-6
Soft 15 (A,4): double against dealer 4-6
Soft 14 (A,3): double against dealer 5-6
Soft 13 (A,2): double against dealer 5-6
All other hands: do not double.
Doubling down means placing an additional bet equal to your original wager and receiving exactly one more card — no hitting after that, no matter what. In exchange for the restriction, you get to invest more money into hands where the math favors you. It's a controlled way to increase your edge, and getting the doubling decisions right is one of the biggest steps between casual play and mathematically sound play.
This page covers every doubling situation in standard blackjack: the hard hand rules, the soft hand rules, why each rule works, the rule variations that change your options (some casinos only let you double on 9, 10, or 11), and the common mistakes that cost the average player 0.2-0.5% in effective house edge. Everything here is based on standard basic strategy for 6-deck blackjack with dealer stands on soft 17 (S17). See the strategy chart for the complete reference.
The mechanics: after receiving your first two cards, you place an additional bet equal to your original wager. The dealer gives you exactly one more card, then you're done — no more decisions, no more cards. If you win, you win double your original bet. If you lose, you lose double.
The strategic value: doubling lets you invest more money into hands where you have a mathematical advantage. If you have 11 against dealer 6, you're in one of the strongest situations in the game — the dealer will bust often, and any card you draw except an ace makes 12-21. Doubling that hand captures the higher expected return by putting more money on the table when the odds are in your favor.
The restriction: you can't hit after doubling. If you double on 9 and draw a 2 for a total of 11, you're stuck with 11. This restriction is why doubling on marginal hands (soft 13 vs dealer 2, for example) doesn't work — you're trading the flexibility to hit again for a bet that isn't actually favorable enough to justify it.
An 11 is the strongest starting hand in blackjack — any 10-value card makes 21, and there are more 10-value cards in the deck than any other rank (four times as many, counting T, J, Q, K). Against dealer up-cards 2 through 10, you have a positive expected return and doubling captures more of it.
Math: With 6 decks, roughly 30.8% of remaining cards are 10-value, giving you close to 30% chance of hitting 21 on your one hit card. Against dealer 6 specifically, doubling on 11 has an EV of about +65% of your original bet — one of the strongest positions in the game.This one depends on the rule set. If the dealer hits soft 17 (H17), double 11 vs A because the dealer's soft-hand hitting means more bust opportunities. If the dealer stands on soft 17 (S17), the recommended play depends on if surrender is available — usually just hit. Most modern strategy charts have this exception marked.
Rule impact: Miscoding this specific situation (11 vs A) is one of the most common basic strategy errors. Check your table's soft-17 rule — it's usually posted on the table felt.A 10 is nearly as strong as 11 — you'll hit 20 with any 10-value draw, and any card from 2 through 9 gives you a competitive total. Against dealer 2 through 9, you have a positive EV and doubling captures it. Against dealer 10 or A, don't double — the dealer's upcard is too strong.
Why not vs 10: Dealer 10 has a strong probability of making 20+, and doubling your bet on a hand that could easily lose to their 20 doesn't make mathematical sense. Just hit — you can still improve to 20 without doubling the risk.Nine is more borderline. Against dealer 3-6 (the "weak" upcards where dealer must draw and often busts), doubling is correct. Against dealer 2 (also weak but slightly stronger than 3-6) — hit, don't double. Against dealer 7+ — hit, don't double.
Why the tight range: With 9 you don't have as much upside as with 10 or 11 — any 10-value card makes 19, which is good but beatable by dealer 20 or 21. Doubling only pays when dealer bust probability is high enough — dealer 3-6 are the sweet spot.Soft hands are hands with an ace counted as 11 (can't bust with a single hit). These behave differently from hard hands because you can't bust — the worst case with a soft 13 taking a hit is that you get 22 with the ace becoming 1, giving you 12. The downside is limited; the upside includes making 17-21 with the right draw.
Soft doubling captures this asymmetric return: you're investing extra money when the downside is capped and the upside is real. The doubling rules for soft hands look counterintuitive at first (double A,2 which is only worth 3 or 13?) but the math works.
Soft 18 is a decent hand — 18 wins if dealer has 17, ties if dealer has 18, loses otherwise. Doubling against dealer 3-6 (weak upcards likely to bust) works because you often stay at 18 while dealer busts, or draw an ace or low card that keeps you at 18-19.
Soft 17 alone is a losing hand — 17 ties dealer 17, loses to dealer 18-21. Standing on it is worse than hitting. Against weak dealer upcards where the dealer often busts, doubling captures both the bust opportunity AND the chance to improve to a stronger total.
These are marginal starting hands (13-16 effective value). Doubling only works against the weakest dealer upcards (5-6 for the weakest soft hands, 4-6 for slightly stronger soft hands). The math is tight — off by one, and hitting is correct instead.
The exact ranges: A,4 and A,5 double vs 4-6. A,2 and A,3 double vs 5-6. Everything else, hit.Reference chart for standard 6-deck S17 DAS rules:
| Your hand | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | A |
|---|---|---|---|---|---|---|---|---|---|---|
| Hard 11 | D | D | D | D | D | D | D | D | D | H* |
| Hard 10 | D | D | D | D | D | D | D | D | H | H |
| Hard 9 | H | D | D | D | D | H | H | H | H | H |
| Soft 18 (A,7) | S | D | D | D | D | S | S | H | H | H |
| Soft 17 (A,6) | H | D | D | D | D | H | H | H | H | H |
| Soft 16 (A,5) | H | H | D | D | D | H | H | H | H | H |
| Soft 15 (A,4) | H | H | D | D | D | H | H | H | H | H |
| Soft 14 (A,3) | H | H | H | D | D | H | H | H | H | H |
| Soft 13 (A,2) | H | H | H | D | D | H | H | H | H | H |
Key: D = Double, H = Hit, S = Stand. *11 vs A: Double if dealer hits soft 17 (H17); hit if dealer stands (S17). All other hands not listed: no double (just hit, stand, or split per basic strategy).
Why is 11 vs 6 such a strong double? Let's work through the actual expected value math.
Just hitting (no double):
Doubling:
The gap: Doubling generates roughly 2× the expected profit because you're risking 2× the money on a favorable hand. The dealer's 6 is the weakest upcard (highest bust probability), your 11 is the strongest starting total, and the math strongly favors putting more money on the table.
Numbers from combinatorial analysis assuming 6-deck S17 DAS. Actual EV varies slightly by exact rule set. See Peter Griffin's The Theory of Blackjack for the underlying probability math, or the house edge article for how these calculations aggregate.
Not every blackjack game lets you double freely. Common restrictions:
Many European tables restrict doubling to hard totals of 9, 10, or 11 only. This removes all soft doubles — soft 18 vs 6, soft 17 vs 5, etc., become "just hit" plays. The house edge impact is about +0.09%, meaningful but not devastating.
Some restrictive tables allow doubling on 10 or 11 only. This removes the 9 vs 3-6 double along with all soft doubles. Larger house edge impact — approximately +0.14%.
If you split a pair, you can't double the resulting hands. Impact: +0.14% to house edge. Common on some Vegas and European tables.
Standard rule on most US casinos — double allowed on any starting 2-card total. This is the baseline for the 0.5% house edge figure. Any restriction increases the edge against the player.
Always check the table rules before playing. The doubling policy is usually printed on the table felt or displayed in the online game info. If you can't find it, ask the dealer or check the game's help screen.
Doubling means putting more money on the table. Some players hesitate because the doubled bet feels psychologically harder to lose. But 11 vs a weak dealer upcard is one of the strongest positions in the game. Refusing to double there is refusing free money the math has already handed you.
Cost of skipping double on 11 vs 6: about −$0.32 per hand at $10 bet sizeSome players see "10 is a strong hand" and double against any dealer upcard. Against dealer 10 or A, don't — the dealer has a strong probability of finishing with 20+, and you're doubling your bet on a hand that's losing more often than winning in that situation.
Cost of double on 10 vs 10: about −$0.14 per hand at $10 bet sizeNever double on 12+. Any 10-value card busts you. The one card you're allowed to draw could immediately end the hand. Basic strategy says stand or hit these hands based on dealer upcard; doubling them is always wrong.
Cost of double on 12 vs 6: about −$0.40+ per hand at $10 bet sizeDoubling A,2 vs 6 feels counterintuitive — you're turning a hand worth 3 or 13 into a mandatory one-card draw. But the math works: dealer 6 busts about 40% of the time, and any 8, 9, or 10 draw gives you a competitive 11-21 total. Skipping soft doubles is a common recreational-player mistake worth about 0.05-0.1% in effective edge over full play.
Cost of skipping soft doubles: about 0.05-0.1% over full playThe opposite mistake: doubling soft 18 vs dealer 9, or soft 15 vs dealer 3. In these situations, the dealer's upcard is too strong or the soft hand isn't weak enough to gain from the doubling structure. Follow the chart — soft doubles are situation-specific.
Cost varies by situation; often −$0.10 to −$0.30 per mistake at $10 bet sizeThe dealer 2 is a weak upcard, but not as weak as 3-6. On 9 vs 2, hitting is correct — doubling adds bet size to a hand that's only marginally favorable. Small mistake but adds up over many hands.
Cost of double on 9 vs 2: about −$0.03 per hand at $10 bet sizeSome players understand the doubling rules mathematically but resist them at the table. The reasons are almost always psychological:
The mathematically correct approach: every double is independent, and each individual doubling opportunity is positive EV or it isn't. Recent results don't change the math on the current hand. Session bankroll should be sized to absorb doubling variance (see our bankroll management).
If you're playing to entertain yourself and can't stomach the doubling variance, that's fine — but understand you're paying about 0.2-0.3% extra in effective edge for the emotional comfort. That translates to real money over many sessions.
Almost always — against dealer 2 through 10, always double. The one exception is 11 vs dealer Ace: double if the dealer hits soft 17 (H17), hit if the dealer stands (S17). Check your table's soft-17 rule.
Standard rules: no. Doubling is only allowed on your initial 2-card hand. If you hit and draw to a 10 or 11, you're past the doubling opportunity. Some casinos allow "double on any total" — check the table rules.
Depends on the rules. Most modern US tables allow "Double After Split" (DAS). Some European and older US tables don't. When DAS is allowed, doubling on 10s and 11s after splitting follows the same rules as normal doubling.
Then you shouldn't be playing at that stake level — your bankroll is too small for the game. Every hand where you can't follow correct basic strategy costs real EV. If $10 tables force you to skip doubles, drop to $5 tables where you can play correctly, or don't play until bankroll is sufficient.
Some casinos allow "double for less" — doubling with a smaller additional bet. It's mathematically worse than a full double: you're accepting the doubling restriction (one card only) without capturing the full EV benefit. If the situation calls for a double, do a full double or don't double at all.
You're stuck with whatever total you have — no additional hits. That's the doubling restriction. If you double on 11 and draw a 2 for a 13, you have to stand on 13 regardless of what the dealer shows. This is part of the doubling calculation; the math already accounts for the times you draw poorly.
Marginally. Basic strategy for doubling is nearly identical for 1-deck through 8-deck games. There are a few borderline situations (like soft 18 vs 2) that flip between hit and double depending on deck count, but 99% of the doubling decisions are the same across all deck counts.
No. Basic strategy accounts for every doubling situation mathematically. Deviations from basic strategy cost expected value. If you want to add strategic complexity, learn card counting — that provides mathematically-grounded deviation triggers. Hunches don't.