Blackjack Basic Strategy Chart

The mathematically optimal action for every hand against every dealer up-card, computed exactly via recursive expected-value analysis (not simulated) under an infinite-shoe approximation — the same assumption virtually every published basic strategy chart uses.

Dealer Stands on Soft 17, DAS, pays 3:2

Why there is no deck-count control: a basic-strategy chart is total-dependent and very nearly deck-agnostic — the same reason published charts do not ship one per shoe size. These cells are computed against an infinite shoe, which is the standard convention and what makes them comparable to any chart you already own. Deck count moves the house edge rather than the plays, and that is where this site varies it: rule impact quotes every edge against a real six-deck shoe dealt without replacement.

SStandHHitDDoublePSplit

Hard totals

Your hand2345678910A
Hard 4
Hard 5
Hard 6
Hard 7
Hard 8
Hard 9
Hard 10
Hard 11
Hard 12
Hard 13
Hard 14
Hard 15
Hard 16
Hard 17
Hard 18
Hard 19
Hard 20

Soft totals

Your hand2345678910A
Soft 13
Soft 14
Soft 15
Soft 16
Soft 17
Soft 18
Soft 19
Soft 20

Pairs

Your hand2345678910A
Pair of Aces
Pair of 2s
Pair of 3s
Pair of 4s
Pair of 5s
Pair of 6s
Pair of 7s
Pair of 8s
Pair of 9s
Pair of 10s

Reading the grid: rows are your hand (hard totals, soft totals with an ace, and pairs), columns are the dealer’s up-card. A hard 12–16 is a “stiff” hand — the chart is where it earns that name, since whether to hit or stand flips by dealer card. Pairs of 8s split against almost everything, even a dealer 10, because two hands starting at 8 beat one hand stuck at a hard 16. Soft hands (an ace counted as 11) double more aggressively than the equivalent hard total, since busting isn’t possible on the next card.

Never take insurance

When the dealer shows an ace, they’ll offer insurance — a side bet, up to half your wager, that pays 2:1 if their hole card is a ten. It sounds like protection; it’s actually a house edge of about 7.7% on its own, worse than almost anything else at the table. Only 4 of the 13 unseen card ranks are worth ten, so the true odds (4/13 ≈ 30.8%) are worse than the 1-in-3 the 2:1 payout would need to break even — the bet loses about 1/13 of a unit on average. That holds regardless of what your own hand is worth, including the tempting-sounding “even money” offer on your own blackjack, which is the identical bad bet under a friendlier name.

Or answer one hand at a time

Three taps instead of parsing the full grid.

What’s my play right now?

Tap your two cards and the dealer’s up-card for an instant answer.

Your first card
Your second card
Dealer shows
Split
8+8 vs dealer 10
Why? See the math →

Try next

Split EV uses the standard infinite-deck approximation (each resulting hand treated independently); resplitting further hands isn’t modeled, which can very slightly understate splitting a few borderline pairs (e.g. 8s against a strong dealer card).

Odds The Law uses cookies for analytics and to show ads. Decline and neither will load. See our Privacy Policy for details.