Playing Against a Dealer 7, Hand by Hand

Every strategy chart has a seam, and in blackjack it runs between the 6 column and the 7 column. Against a 6 the chart is at its most forgiving: stiffs stand, soft hands double, most pairs split. The 7 takes nearly all of that away. It is the first card where every stiff hand hits, where soft doubling, which runs through the 6 column, stops for good, and where the pair rows thin out, and the strategy against it is close to a different game.

The number behind the seam, per the final total distribution tables computed by Wizard of Odds for six decks with the dealer standing on soft 17: a dealer showing a 7 busts 26.2 percent of the time, against 42.3 percent for a 6. The 7 is the first upcard that reaches a standing total with a single ten-value card, draining the bust column and piling probability onto exactly 17, the most likely single finish at 36.9 percent. Our piece on the dealer's final total distribution lays out the full tables; this article walks the 7 column cell by cell, continuing our walks through the 2 column and the 6 column.

Table conditions throughout, matching the app's standard shoe chart: dealer peeks, blackjack pays 3:2, double on any two cards, late surrender where offered, four to eight decks, dealer stands on soft 17 unless stated.

The whole column, one table

One note: nothing surrenders against a 7 in any chart, at any deck count, under either soft 17 rule, because every hand here is worth playing out.

Why every stiff hits

A stiff is a hard 12 through 16, and against a 7 all five totals hit. A stiff is below every total the dealer is allowed to stop on, so standing wins exactly when the dealer busts, which against a 7 happens 26.2 percent of the time, about one finish in four. The other 73.8 percent of dealer outcomes are 17 or higher and beat the stiff outright. Against a 6 the same stand wins 42.3 percent of the time, enough that standing loses less than hitting there.

Hitting is not safe either: a hard 12 busts on the four ten-value ranks and a hard 16 on eight of thirteen. But hitting adds outcomes standing does not have, since a draw that lands on 17 through 21 can beat or tie a dealer who does not bust, and across many trials the hit loses less. The boundary inside the column shows the arithmetic from both sides: a hard 17 pushes the most likely finish and wins the bust, 36.9 plus 26.2, so it wins or pushes 63.1 percent and stands, while a hard 16 loses to that 17 and everything above and wins only the 26.2 bust, so it draws. Our guide to when to hit and when to stand works the same logic across the whole chart.

Two hands that stand: soft 18 and 9,9

Soft 18 stands against a 7, and so does a pair of 9s, and both stand for the same arithmetic. Take an 18 against the dealer's row of outcomes. It wins when the dealer finishes on 17 or busts, 36.9 plus 26.2 for 63.1 percent, pushes the 18 for another 13.8, and loses to 19, 20 and 21 combined, 7.8 plus 7.9 plus 7.4 for 23.1 percent. A hand that wins almost two thirds of the time and ties a seventh more is not one to take apart or redraw.

Below it, soft 13 through 17 all hit: none can bust, and none beats the 17 the dealer makes most often. Soft 18 is the first total that does real work here, as our piece on soft 18 frames it: drawing risks converting a winning hand into a hard 14 or 15. Above it, soft 19 loses only to a dealer 20 or 21, 7.9 plus 7.4 for 15.3 percent, and soft 20 only to the 21, 7.4 percent, so both stand.

The pair of 9s is the same decision wearing a costume. Held together it is an 18, and our guide to when to split 9s gives the answer in one sentence: a dealer showing 7 makes 17 more often than any other total, and your 18 beats 17. Splitting would leave a winning position for two hands that each build from a 9. One upcard later the logic flips: against an 8 the dealer's most likely finish is 18, which ties, and against a 9 it is 19, which beats it, so 9,9 splits again.

The doubles that survive, and the one that does not

Only two hard totals double against a 7: hard 10 and hard 11, the strongest doubling starts in the game. The card most likely to arrive is a ten-value, sixteen of every fifty-two cards, which turns a 10 into a 20 and an 11 into a 21. A doubled 20 loses only to the dealer's 21, 7.4 percent per the table, and a doubled 21 cannot lose, since a dealer showing a 7 can never hold a blackjack and at best ties it with a drawn 21.

Hard 9 is the near miss. In the shoe charts, 9 doubles against a 3, 4, 5 and 6 and hits everything else, and our piece on doubling hard 9 gives the reason in a line: against a 7 and above, the dealer finishes strong too often, and the plain hit keeps the flexibility the double gives away. The 5,5 row belongs here: fives never split, and 5,5 is simply a hard 10 that doubles.

21 TrainerFree on the App Store. Drill the 7 column with instant feedback and the reason on every hand. Get the app

The pairs: what still comes apart

Five pair rows split against a 7: 2,2, 3,3, 7,7, 8,8 and A,A. The 2s and 3s split against a 7 in every chart. The sevens deserve the closest look, since the 7 upcard is the edge of their range: as our piece on splitting 7s explains, a 7 with a ten underneath makes 17, and 17 is a total the split hands can beat or push. You are splitting into the dealer's worst made hand.

The contrast with 6,6 in the same column teaches the lesson from the other side. As our article on splitting 4s and 6s puts it, a split 7 that catches a ten-value card makes 17, while a split 6 that catches one makes 16, still a stiff. Two hands from 6 against a dealer who finishes on a made hand 73.8 percent of the time is two likely losers, so the 6s hit the 12. The 4s hit too: 4,4 is a hard 8, and its split case depends on doubling after a split against weaker upcards.

The famous rows behave as always. A pair of 8s is a hard 16, the worst total in the game, and splitting aces and eights holds against every upcard: two ordinary hands from 8 lose less than one hopeless 16. A pair of aces is a soft 12 with almost no upside, while each split ace starts from the best position in the game. And 10,10 stands: a made 20 that loses only to the 7.4 percent dealer 21 is not worth trading for two hands starting over.

How the rules move this column, and how one cannot

Deck count and the soft 17 rule move borderline cells elsewhere in the chart, but in this column the soft 17 rule does nothing, at any deck count, and the reason is arithmetic. A soft 17 is an ace counted as 11 plus exactly 6 more points, and a dealer showing a 7 already holds 7 points before any ace enters the picture, more than those 6, so that dealer can never hold a soft 17. The rule bites only on the ace and 2 through 6 columns, which is why our S17 versus H17 comparison lists its changes there and nowhere near this column.

Deck count moves exactly one cell. In four to eight deck charts, 6,6 against a 7 is always a hit; in the app's one deck and two deck charts it becomes a split when doubling after a split is allowed, otherwise a hit. Every other cell is identical across all six of the app's charts: learn this column once and it carries across rule sets.