Blackjack
Blackjack Basic Strategy Calculator
Inputs
Results
Recommended Action
Hit
Pair Mode
Off
Rule Set
6D S17 DAS
How to use Blackjack Basic Strategy Calculator
Leave the form on its shipped defaults: Hand Total 16, Dealer Upcard 10, Pair Rank set to Not a pair, Soft Total unchecked. That is a hard sixteen against a ten-value upcard, the stiff that generates more floor arguments than any other two-card total. Submit the fields. The engine returns Hit as the recommended action, Pair Mode Off, and Rule Set 6D S17 DAS. Standing a hard 16 versus 10 loses whenever the dealer’s hole card completes 17 through 21, which is frequent. Hitting busts often, yet the expected loss of the hit is still smaller than the expected loss of the freeze. The label is a ranking of two bad options, not a claim that a five will appear.
Change only Pair Rank to 8 while keeping Dealer Upcard at 10. The numeric total is still sixteen, but pair logic runs first. Eights always split, so the action flips from Hit to Split. A pair of fives is not played as a stiff: the engine treats 5,5 as a ten for doubling, so versus a six it would Double and versus an ace it would Hit. Tens stand. Aces split. Those overrides exist so you never treat 8,8 as a hard sixteen or 5,5 as a ten that must be hit like a stiff.
Leave Soft Total false for the default 16. Checking it reinterprets the same number as Ace-5. Soft sixteen versus ten is a hit; soft doubles on 15–16 exist only against 4–6 under this chart. Soft 18 versus 9, 10, or ace is a hit; versus 3–6 it is a double; versus 2, 7, or 8 it stands. Mixing the hard and soft grids is the most common lookup error at a six-deck shoe.
A second walk-through with adjacent numbers: Hand Total 12, Dealer Upcard 3, no pair, not soft. The hard chart hits 12 versus 2 or 3 and stands only versus 4–6. Hard 13–16 stand against 2–6 and hit against 7–ace. Hard 17 and above stand. Hard 11 doubles against every upcard except an ace; hard 10 doubles against 2–9; hard 9 doubles against 3–6. When those totals are entered here, the same cells fire. The tool does not ask for suits, composition, or count.
The printed rule set is the contract. Six decks, dealer stands on soft 17, double after split allowed. If the table is H17, no DAS, or single-deck, some doubles and splits move. Mapping a live hand onto this grid is educational, not a plus-EV conversion.
About this calculator
Basic strategy is a complete map of the action that minimises expected loss for every two-card total against every dealer upcard under a stated rule set. It is not a winning system. Under 6D S17 DAS the house still keeps a fraction of a percent against a player who never departs from the chart. The purpose of this calculator is to look up that chart without memorising 200-plus cells, and to make the pair-versus-hard-versus-soft priority explicit so a sixteen of eights is not played as a stiff.
The first published computer-assisted tables came from Roger Baldwin, Wilbert Cantey, Herbert Maisel, and James McDermott in the Journal of the American Statistical Association in 1956. They used desk calculators at Aberdeen Proving Ground, not a casino mainframe. Edward O. Thorp’s Beat the Dealer in 1962 popularised both the strategy and the idea that the remaining composition can shift the edge. Julian Braun at IBM later refined multi-deck indices that casinos still approximate on the plastic cards they sell in gift shops. This tool implements a compact 6D S17 DAS subset: six decks, dealer stands on soft 17, doubling after a split permitted.
Limits are structural. The chart assumes the dealer’s hole card is unknown, the shoe is randomly ordered, and you may double any two-card total after a split when the double cells fire. It does not model late surrender, restricted doubles (9–11 only), peek versus no-peek, or Spanish 21. It does not count tens. A true count of +3 can flip hard 16 versus 10 from hit to stand; this calculator will still say Hit because it is a zero-count chart. Insurance is a separate side bet and is not offered here.
History on the felt diverged from the paper. Nevada’s 1960s shoes were often four decks; Atlantic City added early-surrender windows that Thorp-era players exploited until the rules were rewritten. Online continuous shufflers reset composition every hand, which makes counting useless and basic strategy the entire skill overlay. Live 6D S17 DAS remains one of the lowest-edge table products still dealt to the public, which is why the label is printed on every result row.
Use the lookup as a training aid and as a check against folklore: never split tens, never stand a soft 17, never invent a third action called “hope.” The calculator will not tell you whether to sit at the table. It ranks the least-bad action for the totals you typed under this rule set.
Math under the hood
Basic strategy is an expected-value maximising policy printed as a table, not a temperament and not a slogan. Roger Baldwin, Wilbert Cantey, Herbert Maisel, and James McDermott published the first complete maximisation in the Journal of the American Statistical Association in 1956, enumerating dealer bust frequencies and final-total frequencies under an infinite-deck approximation, then storing in each cell whichever of Hit, Stand, Double, or Split produced the larger mean settlement. Edward Thorp's 1962 book and Julian Braun's later computer enumerations replaced that infinite-deck sketch with finite-shoe combinatorics. On a six-deck shoe where the dealer stands on soft seventeen and doubling after a split is allowed, the qualitative map is stable across those enumerations.
Construction of a cell is a comparison of means. Against each upcard, every legal action is scored by the expected payoff of the resulting distribution of player totals versus the dealer's completion, including the chance the dealer busts. Soft eighteen versus a six is stored as Double: the extra unit is justified against a bust-prone upcard, and the ace still cushions a small third card so the hand does not explode. Hard sixteen versus a ten is stored as Hit. Standing surrenders too many dealer twenties; the 1956 paper and Braun's later counts keep that cell on the hit side at a zero count. A pair of eights is always split, replacing one miserable sixteen with two live eights. A pair of fives is never split; the two cards are treated as a hard ten, so the double-or-hit policy for ten applies.
Worked defaults on the six-deck, stand-seventeen, double-after-split table recover the following cells. Soft 18 versus 6 is Double. Hard 16 versus 10 is Hit. Eight and eight (8,8) versus any upcard is Split. Five and five (5,5) versus 6 is Double, because the total is a 10; 5,5 versus an ace is Hit. Hard seventeen and above stand. Hard thirteen through sixteen stand versus two through six and hit versus seven through ace. Hard twelve stands only versus four through six. Hard eleven doubles versus every upcard except an ace. Those are the same decisions the 1956 authors would have printed for this rule set.
Assumptions sit under every cell and should be named before anyone treats the table as scripture. Cards are treated as independent apart from the two player cards and the upcard already in play, so composition-dependent exceptions such as twelve versus two with a deuce in hand are averaged out. Blackjack is paid three to two in the modern shoe; the qualitative grid for these particular cells does not flip if one uses even money in the older papers. Surrender is off. Expected-value gaps between adjacent cells are often a few hundredths of a unit, which is why the folklore that always stands sixteen is costly over tens of thousands of hands even though any one hand is drowned in variance. The table reports the policy that maximises the mean, not a dollar figure for a single coup.
History on the felt is a footnote to the arithmetic. Nevada four-deck shoes of the 1960s, Atlantic City early-surrender windows, and continuous shufflers all change how much the remaining pack can drift, but they do not rewrite the zero-count ranking of hard sixteen versus ten or of eight-eight versus anything. Learn the table as a ranking of two bad options when the cell is a stiff, and as a ranking of two good options when the cell is a double. The mean is what Baldwin's group maximised. Variance is what the session still delivers.