When to Double Down in Blackjack

Every doubling situation in blackjack, why each one is correct, and the four doubles that players miss most often. Doubling correctly is worth about 1.6% of your action.

Updated September 2026 · how these numbers are calculated

Doubling down is the largest single source of value in basic strategy after the split decisions. A player who never doubles gives up roughly 1.6% — nearly four times the entire house edge at a good table.

The reason is structural. Every other decision in blackjack is about limiting losses. Doubling is the only one that lets you put more money in when the odds have swung your way.

The complete list

Dealer upcard across the top. Double where marked; otherwise play the hand normally.

4–8 decks, dealer stands on soft 17, double after split allowed.

On A,7 against 3–6, double if the table allows it; if it does not, stand rather than hit. That is the “Ds” notation on the full strategy chart.

Why each group is correct

Hard 11 — the strongest double in the game

You cannot bust. Any ten-value card makes 21, and there are sixteen of them in every fifty-two. You reach 20 or 21 about 46% of the time on the draw card alone.

Double against everything except an ace. Against an ace, the dealer’s hand is strong enough and their natural-blackjack probability high enough that the extra bet stops paying. (In games where the dealer hits soft 17, doubling 11 against an ace becomes correct.)

Hard 10 — nearly as strong

Same logic, one rank lower: a ten-value card gives you 20 rather than 21. Double against 2 through 9. Stop against a ten or an ace, where the dealer is likely to have 20 or better and you are drawing to merely match it.

Hard 9 — narrow window

Nine only doubles against the dealer’s genuinely weak cards, 3 through 6. Against a 2 the dealer is not weak enough; the 2 busts only about 35% of the time, the least bust-prone of the “bust cards”. Against 7 or higher you are the underdog and should just hit.

Soft doubles — a different bet entirely

Hard doubles are about making a strong total. Soft doubles are about punishing a weak dealer. On A,4 against a 5 you are not expecting to make 19. Roughly half your draws leave you with a mediocre hand. You double anyway because the dealer showing a 5 busts about 42% of the time, and getting more money onto the table while they are that vulnerable is worth more than the quality of your own total.

This is why the soft-double windows are narrow and centred on 4, 5 and 6 — the three upcards where the dealer busts most often:

Dealer stands on soft 17. The ace figure assumes the dealer has already checked for and does not have a natural.

Notice that 5 and 6 are the peak, and 6 is slightly below 5 — a detail that surprises people who think of 6 as the dealer’s worst card. It is the worst card in terms of the dealer’s final expected hand, but 5 busts marginally more often.

5,5 — a doubling hand, not a splitting hand

A pair of fives is a hard 10. Play it as a hard 10 and double against 2–9. Splitting fives is one of the most expensive errors available at the table: it converts the best low starting total in the game into two hands starting with the worst card.

The four doubles most often missed

All four share a pattern: they feel like pressing a hand that is not strong enough. Hard 9 does not feel like a doubling hand. Soft 18 feels finished. The chart is not describing how the hand feels; it is describing what the hand is worth against that specific upcard.

When not to double, regardless of the chart

• If the table does not allow it. Doubling on 9 is barred at some tables; the chart assumes it is available. Hit instead.
• If you cannot cover the second bet. Never double with money you need to leave with. Basic strategy assumes an indifferent bankroll; you may not have one.
• After splitting, where DAS is not offered. Some tables bar doubling after a split. Check before you split a pair whose value depends on doubling the resulting hands.

What it is worth

Playing every double correctly, against never doubling at all: about 1.6% of total action. On $10 hands at 60 hands an hour, that is roughly $9.60 an hour — more than the entire house edge at a well-played 3:2 table.

Next: the complete basic strategy chart.