Doubling down

Doubling is the player's best offensive tool, and how freely a table allows it is worth 0.09 percentage points.

What doubling is

You double your stake and receive exactly one more card, then you must stand. It is the only moment in blackjack where you can increase a bet after seeing information, which is why it matters out of proportion to how often it comes up.

The restriction and what it costs

Liberal tables let you double on any two cards. Restrictive tables allow it only on hard 9, 10 and 11, which removes every soft double.

Running the calculation both ways gives +0.088 percentage points for the restriction, taking the edge from 0.515% to 0.603%.

Where the value is

Hard 11 is the strongest double in the game: one card, and four ranks in thirteen make it 21. Hard 10 is close behind. Hard 9 doubles only against a weak dealer upcard.

Soft doubles confuse people because a soft 17 does not look worth doubling. The point is not the total; it is that the hand cannot bust and the dealer is weak. You are pressing a bet against a dealer likely to break, and our dealer upcard odds page shows how likely.

Never double a hard 12 or higher: any ten-valued card busts you and you are forced to stand on one card.

Frequently Asked Questions

What is the best hand to double?
Hard 11. One card, and four ranks in thirteen make 21.
Why double a soft 17?
Because it cannot bust, and against a weak dealer upcard you are pressing a bet the dealer will often lose by busting.
What does the 9-11 restriction cost?
0.09 percentage points, taking the edge from 0.52% to 0.60%.
Can I double after splitting?
It depends on the table. Where it is not allowed it costs 0.13 percentage points.

Every figure on this page is computed by an expected-value engine that plays the game out: it derives the dealer's final-total distribution by recursion, evaluates standing, hitting, doubling and splitting at each decision, and takes the best. Rule costs are measured by re-running the whole calculation with the rule changed, not by quoting a published delta. The model assumes independent card draws, which sits a few hundredths of a percent from an exact shoe analysis; the rule differences it produces match published values. Rules vary by table, so read the felt.