Blackjack basic strategy

What basic strategy is

Basic strategy is the play with the highest expected value for every combination of your hand against the dealer's up card, assuming you know nothing about what is left in the shoe.

It is not a formula or a rule of thumb — it is the result of an exhaustive calculation. The first academic work to derive it was published in 1956 in the Journal of the American Statistical Association by four mathematicians in the US Army, who did the calculation on mechanical calculators.

Two points surprise anyone meeting the table for the first time:

• The dealer's up card matters as much as your hand. The same 16 is played differently against a 6 and against a 10. Most beginners' mistakes come from looking only at their own cards.
• Some plays are about losing less, not winning. In some hands every option is bad, and the table picks the least bad one. That feels counterintuitive, and it is exactly where the value lies.

The rules the table assumes

There is no single "basic strategy". It follows from the table rules, so a chart is valid only for the conditions it was calculated for. The chart on this page is calculated for the most common conditions:

A change in the rules changes cells in the table. Two cases worth knowing:

• A dealer who hits soft 17 (H17) — a rule that is harder on the player, and changes several decisions among the soft hands.
• Blackjack paying 6:5 — it changes no cell in the table, but roughly triples the house edge. It is the one rule worth recognising at once: a table like that wipes out the entire edge counting can produce.

Hard hands

A hard hand holds no ace, or an ace that must count as 1. It is the most common case at the table.

The central logic here is splitting the dealer's cards into two groups:

• Up card 2–6: the dealer is weak. The dealer will have to draw, and the chance of busting is high. Your job is simply not to bust first — so you stand on hands that look weak, like 13 or 14.
• Up card 7–A: the dealer is strong. The dealer will probably finish on 17 or more. A 15 will not beat that, so you hit despite the risk — not because it is safe, but because the alternative is worse.

The famous case is 16 against 10, the worst hand in the game. Basic strategy says to hit, and most players cannot bring themselves to. Both options lose; hitting loses slightly less. It is also the hand where the count changes the decision first.

Soft hands

A soft hand holds an ace counting as 11, so a single card cannot bust it — the ace can always drop back to 1. That changes the decision completely, and it is where players go wrong most.

The classic example: A,7 against a dealer's 9. A soft 18 sounds like a good hand to stand on. But 18 loses to any dealer who finishes on 19 or more, and against a 9 showing that is very likely. Since hitting here costs nothing — you cannot bust — you hit.

The general rule for soft hands: against a weak dealer (3–6) you double a lot, because two advantages combine — the dealer may bust, and you risk nothing.

Pairs and splitting

Two cards of the same rank can be split into two separate hands. Some of the rules here feel wrong and are mathematically right anyway:

• A,A and 8,8 are always split. Even 8,8 against a 10 — not because it is pleasant, but because two hands starting at 8 lose less than one hand of 16.
• 10,10 is never split. A 20 is the second-best hand in the game. Splitting it trades one excellent hand for two middling ones — an expensive mistake, and a common one.
• 5,5 is never split. It is better played as a 10 and doubled.
• 4,4 is split only against 5 and 6 — and that depends directly on the DAS rule. At a table that does not allow doubling after a split, they are not split at all.

The full table

Rows are your hand; columns are the dealer's up card.

• H Hit
• S Stand
• D Double — or hit, if doubling is not allowed
• P Split

Why it comes before counting

This is not a teaching preference but arithmetic:

There is a cognitive side too. At a real table the count runs in parallel with the decision about the hand, both under time pressure. If the decision still takes thought — rebuilding the table, hesitating — it takes the resource the count needs, and the count breaks. So basic strategy has to be automatic before you start counting, not alongside it.

The practical test is simple: a random hand against a random dealer card, the right decision within a second, without rebuilding the table in your head. Until that is true, there is no point adding a count.