Doubling Down: Where Basic Strategy Actually Makes Its Money
Key Takeaway: Doubling is worth about 1.6% to a basic strategy player. Skip it and you hand that back to the house. Hard 9, 10, and 11 account for roughly 94% of that value. When the chart says double, you put the second bet out every time, with no hesitation and no "feel."
Most of what you do at a blackjack table is damage control. You stand on 13 against a 2 because it loses less than hitting. You hit 16 against a 10 because it loses slightly less than standing. Large stretches of basic strategy are about choosing the smaller loss, and that is fine, because those decisions are what keep the house edge near half a percent. They do not make you money.
Doubling is different. It is one of the few places where you get to press. You hold a strong starting total, the dealer is either weak or not strong enough, and the rules let you add a second bet equal to your first. You give up one thing in exchange: you take exactly one more card and nothing after it. Learning when that trade is profitable separates players who lose slowly from players who play at the true theoretical edge.
If you have not yet locked in the full chart, start with our learn basic strategy guide. This article assumes you know the layout and want to understand the reasoning behind the doubling column.
The 1.6% Problem: What Timid Doubling Costs You
Run a simulation of a standard six-deck game with perfect basic strategy, then remove the double option entirely. The house edge rises by roughly 1.6 percentage points. A game that was costing you about 0.5% now costs you around 2.1%. That one change does more damage than switching from 3:2 to 6:5 blackjack payouts, a rule players complain about constantly. Meanwhile many of those same players quietly decline doubles on 9 and 10 against small cards and never notice what it costs them.
You do not need to stop doubling completely to pay this tax. Most recreational players double selectively. They double 11 against a 6 because it feels obvious. They skip 9 against a 3 because "it's only a 9." They skip soft 17 against a 4 because "I already have 17." Every skipped double is a small, real leak, and those leaks add up. At $25 average bets and 80 hands an hour, a 1% leak costs $20 an hour. Over a 200-hour year, that is $4,000 lost simply because you were reluctant to double.
A double makes two changes at once. It doubles your bet, so a positive expectation becomes twice as large in dollars. It also locks you to one card, which destroys the value of being able to hit again. Doubling is correct exactly when the first effect is bigger than the second. Against a dealer 5 or 6, the flexibility you give up is worth almost nothing, because you would stand on any total of 12 or more anyway. Against a dealer 10 or Ace, that flexibility matters a great deal, and the double has to be justified by very high equity alone.
The Math Behind the Decision
The decision rule is simple to state and easy to misapply. Compare two quantities: the expected value of taking one card and standing, multiplied by two, against the expected value of playing the hand out optimally with as many hits as you want. The full derivations across every upcard are laid out in the doubling down EV curves on BlackjackMath.org. Below is the framework, worked through on two benchmark hands using infinite-deck approximations under S17 rules with dealer peek.
Double ⇔ 2 × EV(one card, then stand) > EV(optimal hitting)
EV(one card) = Σ P(card c) × EV_stand(total + c | dealer upcard)
──────────────────────────────────────────────
HARD 11 vs DEALER 6 (dealer bust ≈ 42.3%)
Draw A–5 → 12–16 : EV_stand ≈ −0.154 (5/13 weight)
Draw 6 → 17 : EV_stand ≈ +0.012
Draw 7 → 18 : EV_stand ≈ +0.284
Draw 8 → 19 : EV_stand ≈ +0.490
Draw 9 → 20 : EV_stand ≈ +0.704
Draw T → 21 : EV_stand ≈ +0.903 (4/13 weight)
EV(one card) ≈ 4.332 / 13 ≈ +0.333
EV(hit) ≈ +0.333 (you stand on 12+ vs 6 anyway)
EV(double) ≈ +0.667 → doubling exactly doubles your profit
──────────────────────────────────────────────
HARD 11 vs DEALER 10 (dealer bust ≈ 21.2% after peek)
EV(one card) ≈ +0.058
EV(hit) ≈ +0.076 (re-hitting 12–16 recovers ~0.018)
EV(double) ≈ +0.117
Edge of double over hit ≈ +0.04 units → still a clear double
Look at what the first example shows. Against a 6, the option to hit again has zero value on hard 11, because no card leaves you with a total you would want to hit. Doubling costs you nothing and doubles your profit. Against a 10, re-hitting is worth about 0.018 units, but the doubled bet still comes out about 0.04 units ahead. Move the dealer up to an Ace in a six-deck S17 game and the margin falls to essentially zero, which is why that cell changes with the rules. More on that below.
Hard 11: The Titan of Doubling
Hard 11 is the strongest doubling hand in the game. About 4 in 13 draws give you 21, and you never bust. It is the one doubling hand where you double into strength. You double against a 10 even though that 10 makes 20 about a third of the time.
- Six-deck or eight-deck, S17: Double against 2 through 10. Hit against an Ace.
- Six-deck or eight-deck, H17: Double against 2 through Ace. When the dealer must hit soft 17, the Ace upcard is somewhat weaker, because the dealer turns some soft 17s into busts or weak totals. That is enough to tip the decision to a double. Our dealer soft 17 rule breakdown covers the full set of H17 changes.
- Single and double deck: Double against the Ace regardless of the soft-17 rule. Removing your own two small cards from a short shoe makes the remaining deck richer in tens, which helps your 11.
The 11-versus-Ace decision is the closest call in this group. Missing it costs you only a small fraction of a unit per occurrence. Missing 11 against a 5 or 6 costs a third of a unit every time. Get the large decisions right automatically, then refine the close ones.
Hard 10: Double Against Everything Except 10 and Ace
Hard 10 follows the same logic with a little less margin. You still catch a ten-value card about 31% of the time, which gives you 20. Drawing an Ace gives you 21. You cannot bust. The rule is the same in every common game:
- Double against 2 through 9.
- Hit against 10 and Ace.
Against a 10, your best one-card outcome is 20, which only pushes the dealer's most likely total. The doubled bet does not have enough equity to make up for losing the option to hit. The 10-versus-9 double is the one players most often skip. Many players think of a dealer 9 as "strong." In fact, a dealer 9 makes 19 most often. Your 10 makes 20 or 21 about 38% of the time and beats a 19 from 20 alone. Double it.
Hard 9: The Most Neglected Double
Hard 9 is where recreational players lose the most money through hesitation. A one-card 9 hits 19 on a ten-value card, and 19 does not feel strong. The case for doubling rests almost entirely on the dealer's bust probability. Approximate S17 bust rates by upcard are: 2 ≈ 35%, 3 ≈ 37%, 4 ≈ 40%, 5 ≈ 43%, 6 ≈ 42%. Once the dealer bust rate clears roughly 37%, a moderate total doubled is worth more than the same total played flexibly.
- Multi-deck: Double against 3 through 6. Hit against 2 and 7 through Ace.
- Single and double deck: Add the dealer 2. Double against 2 through 6.
The margins here are thinner than on 10 or 11. They are still positive, and they come up often. A 9 shows up as a starting total frequently enough that skipping these doubles is a steady, measurable leak.
Benchmark Table: Hard Doubling Across Game Types
| Player Hand | 6D/8D, S17 | 6D/8D, H17 | Single/Double Deck | Approx. Dealer Bust Range Targeted |
|---|---|---|---|---|
| Hard 11 | Double 2–10 | Double 2–A | Double 2–A | 17%–43% (all upcards) |
| Hard 10 | Double 2–9 | Double 2–9 | Double 2–9 | 23%–43% |
| Hard 9 | Double 3–6 | Double 3–6 | Double 2–6 | 35%–43% |
| Hard 8 | Never | Never | Double 5–6 (single deck only) | 42%–43% |
To generate a chart that matches your exact rule set, including deck count, soft-17 rule, and double-after-split, use the Strategy Chart Builder. Do not memorize a six-deck chart and then sit down at a pitch game expecting it to fit.
Soft Doubling: Small Edges, Real Money
Soft doubles work on different logic. You are not doubling because your hand is strong. Soft 13 is a bad hand. You are doubling because a soft total cannot bust on one card, while the dealer's 4, 5, or 6 busts more than 40% of the time. The double puts more money in play in the situations where the dealer is most likely to beat themselves.
A useful shorthand is "soft doubles live against 4 through 6." The exact multi-deck pattern is more specific:
- A,2 and A,3: Double against 5–6. Otherwise hit.
- A,4 and A,5: Double against 4–6. Otherwise hit.
- A,6: Double against 3–6. Otherwise hit. Never stand on soft 17.
- A,7: Double against 3–6 in S17, and against 2–6 in H17. If doubling is not allowed, stand. Stand against 7 and 8. Hit against 9, 10, and Ace.
- A,8: Double against 6 in H17 games only. If doubling is not allowed, stand. Stand everywhere in S17.
The pattern widens as the soft total rises. A higher soft total gives the one-card draw a better floor, so the double becomes profitable against a slightly stronger upcard. Our soft hands guide covers the full soft matrix, including the A,7 three-way decision that trips up more players than any other cell on the chart.
Sharp Tip (Marcus Vance): At a rule-restricted table where you cannot double soft hands, remember the "double, otherwise stand" cells. A,7 against 3–6 becomes a stand, not a hit. A,8 against 6 under H17 becomes a stand. Every other soft double defaults to a hit. Most mistakes I see on restricted tables come from players hitting soft 18 against a 5 because they "wanted to double anyway."
Casino Doubling Restrictions: Pricing the Rule Card
Before you sit down, read the felt and the rule placard. Doubling restrictions are among the most common ways a casino quietly raises the house edge. Players rarely price them correctly because the numbers look small. The figures below are approximate multi-deck costs relative to unrestricted doubling on any two cards (DA2).
| Doubling Rule | Approx. Added House Edge | Share of Doubling Value Retained | What You Lose |
|---|---|---|---|
| Any two cards (DA2) | Baseline (0.00%) | 100% | Nothing |
| Hard 9–11 only | +0.09% | ≈94% | All soft doubles; hard 8 in single deck |
| Hard 10–11 only | +0.18% | ≈89% | Soft doubles plus all hard 9 doubles |
| No double after split (NDAS) | +0.14% | N/A (separate rule) | Doubles on post-split 10s and 11s; also weakens several pair splits |
| No doubling at all | ≈+1.6% | 0% | The primary profit engine of basic strategy |
The third column explains why this article focuses on 9, 10, and 11. If restricting doubles to hard 9–11 costs only 0.09%, and doubling as a whole is worth about 1.6%, then those three totals carry roughly 1.51 / 1.60 ≈ 94% of the value. Soft doubles and single-deck hard 8 are real money, but they are the remainder. You should still take them. Just recognize that a player who doubles soft hands perfectly but skips 9 against a 4 has the priorities reversed.
Volatility: Why Doubling Feels Worse Than It Is
Doubling has a price beyond expected value, and that price is variance. Basic strategy blackjack runs a standard deviation of about 1.15 units per hand. Doubles and splits account for a large share of that figure. Counting doubles and splits, your average money in play per round runs about 13% above your initial bet. Roughly one hand in ten is a double.
Doubles that lose tend to cluster in memory. Suppose a typical hard double loses about 40% of the time. The probability of losing three in a row is about 0.4³ = 6.4%, so you will see it roughly once every sixteen strings of three doubles. At a 2-unit loss each, that run costs 6 units on hands where you "did everything right." Players who size their bets aggressively relative to their session bankroll feel this hard, and some respond by doubling less. That is the worst possible reaction, because it cuts the expectation while keeping most of the variance.
The correct adjustment is bet sizing, not strategy deviation. If a string of lost doubles threatens your session, your base unit is too large. As a working floor, carry at least 40 base units for a session and several hundred for a longer trip. Then you can absorb doubling swings without it affecting your decisions. Your bankroll should cover the full strategy. The strategy should never be trimmed to fit your bankroll.
Beating Loss Aversion at the Table
Behavioral research consistently finds that people weigh a loss at roughly twice the psychological weight of an equal gain. At the table, this shows up as a specific bias. Losing 2 units on a double feels much worse than losing 1 unit on a hit, even when the double has the higher expected value. Your mind compares the bad outcome against the alternative. The math compares averages across thousands of trials.
A few practical habits help:
- Decide before the cards come out. Know your doubling triggers so well that the decision is made when the dealer's upcard lands. Hesitation creates room for second-guessing.
- Judge decisions, not results. Doubling 11 against a 6 and catching a deuce is a correct play with a bad result. Hitting it and catching a ten is an incorrect play with a good result. Only the decision is under your control.
- Think in occurrences. Each 11 against a 6 is worth about a third of a unit more when doubled than when hit. A hundred of those over a year is about 33 units. The pain of any single lost double is noise inside that number.
- Always double for the full amount. Most casinos allow you to "double for less." Every dollar you hold back on a correct double is expected value you give away voluntarily. If the full double makes you uncomfortable, the problem is your base bet size.
The Bottom Line
Doubling down is where basic strategy stops defending and starts winning. Hard 11 is a near-automatic double, with the Ace as the only cell that depends on the rules. Hard 10 doubles against everything but 10 and Ace. Hard 9 doubles against 3 through 6 in a shoe and against 2 through 6 in pitch games. Soft doubles concentrate against 4 through 6, widening to 3 through 6 for A,6 and A,7. Read the rule card, price the restrictions, size your bets so variance never changes your decisions, and put the second bet out every time the math calls for it. That is how you collect the 1.6% that basic strategy offers.
