# Drilling the Dealer Upcard Until It Reads Itself

A drill set for reading dealer upcards as families instead of faces: weak 2 through 6, strong 7 through ace, how many chart rows change at each step, and how to practice.

Source: https://21trainer.app/articles/drilling-dealer-upcard-reads/
Updated: October 8, 2026
Publisher: 21 Trainer, makers of Blackjack Strategy: 21 Trainer, an educational blackjack trainer for iPhone and Android (App Store: https://apps.apple.com/app/blackjack-strategy-21-trainer/id6759527695, Google Play: https://play.google.com/store/apps/details?id=app.blackjacktrainer)

> **Educational content, not gambling advice.** 21 Trainer and this article teach blackjack strategy and card counting as skills, using virtual chips only. No strategy or counting system guarantees winnings, and nothing here encourages real-money play. If gambling is a problem for you or someone you know, call 1-800-522-4700 (National Council on Problem Gambling).

Every decision in basic strategy has two inputs: your hand and the dealer's upcard. Many learners drill the first and squint at the second. They count their total carefully, then glance at the dealer's card, run a little lookup, and answer. Fluent readers do not see letters, and fluent players do not see "a 6". They see weak, and the play that follows falls out of that read.

This article is about training that first half until it costs nothing. The material is small, ten upcards and two families, and the payoff is large: the upcard read runs on every hand, and it is a step practice routines rarely isolate.

The chart used throughout is the app's four to eight deck chart with the dealer standing on soft 17: 28 rows, hard 8 through 17, soft 13 through 20, and the pairs 2,2 through A,A. Every count below was taken from that chart.

## Two families and one seam

The organizing fact is the dealer's bust rate, laid out in our guide to [reading the dealer upcard](https://21trainer.app/articles/dealer-upcard-guide/) and the [final total distribution tables](https://21trainer.app/articles/dealer-final-total-distribution/). Against the weak upcards, 2 through 6, the dealer breaks often: per the six deck stands on soft 17 tables, a 6 busts about 42.3 percent of the time and a 5 about 41.8. Against the strong upcards, 7 through ace, the bust rate collapses; a 7 busts about 26.2 percent, and the rate stays below that through the ace. A 7 is the first upcard that reaches a standing total with a single ten-value card.

Basic strategy tracks that cliff almost perfectly. The weak family is where the chart gets aggressive: stiffs stand, soft hands double, most pairs split. The strong family is where it turns defensive: stiffs hit, doubles thin out, and the surrenders appear. The seam between the 6 and the 7 is the biggest boundary on the chart.

How much does it carry? Count the rows whose play differs between neighboring columns. Between the 6 and the 7, 15 of the 28 rows change: hard 9, every stiff total, all six soft totals from 13 through 18, and the 4,4, 6,6 and 9,9 pairs. More than half the chart turns over at that one line, and no other step comes close, which is the case for drilling the family read first: it is the question the chart is mostly answering.

## How much moves at each step

The full count of changing rows between each pair of neighboring upcards, taken from the chart above:

| Upcard step | Rows whose play changes | Which rows |
| --- | --- | --- |
| 2 to 3 | 4 | hard 9, soft 17, soft 18, 6,6 |
| 3 to 4 | 5 | hard 12, soft 15, soft 16, 2,2, 3,3 |
| 4 to 5 | 3 | soft 13, soft 14, 4,4 |
| 5 to 6 | 0 | none |
| 6 to 7 | 15 | hard 9, hard 12 through 16, soft 13 through 18, 4,4, 6,6, 9,9 |
| 7 to 8 | 4 | 2,2, 3,3, 7,7, 9,9 |
| 8 to 9 | 2 | hard 16, soft 18 |
| 9 to 10 | 4 | hard 10, hard 15, 5,5, 9,9 |
| 10 to ace | 2 | hard 11, hard 15 |

Three facts fall out of the table. The 5 to 6 step moves nothing: in this chart the two columns are identical, so a 5 and a 6 can be read as the same card. Inside the families the steps are small: the four weak steps change 12 rows in total and the four strong steps another 12, against 15 at the seam alone. And the two steps that change only two rows, 8 to 9 and 10 to ace, are worth knowing by name: hard 16 and soft 18 at the 9, hard 11 and hard 15 at the ace.

## The second layer inside each family

Once the family read is automatic, a second layer of detail refines it. On the weak side, the 2 and the 3 are the cautious end: the dealer breaks less often against them than against a 5 or 6, so hard 12 hits against both, and hard 9 does not double against the 2. The 5 and 6, as the table shows, are identical here.

On the strong side, the 7 is the mild end: surrenders do not exist against it, and soft 18 still stands. The 8 differs from it only in four pair rows, and the step to the 9 changes only hard 16, which becomes a surrender where offered, and soft 18, which starts hitting. The 9, 10 and ace are where the hardest rows live: hard 16 surrenders against all three in this chart, hard 15 surrenders against the 10, and soft 18 hits against all three. The stiff block tells the same story in one shape: hard 12 through 16 stand against the 4, 5 and 6, 13 through 16 also stand against the 2 and 3 where 12 hits, and everything from the 7 up hits, with those surrenders carved out where the rule exists.

You do not need to memorize the counts table, just the families, the seam, and the boundary rows where a step changes something famous. The published column walks cover them in depth: our pieces on [the 6 column](https://21trainer.app/articles/basic-strategy-vs-dealer-six/) and [the 7 column](https://21trainer.app/articles/basic-strategy-vs-dealer-seven/) are the two halves of the seam, and [the 2 column](https://21trainer.app/articles/basic-strategy-vs-dealer-two/) shows the weak family at its most cautious.

## Four drills for the read

**Call the family before the play.** Take a deck, flip one card at a time, and for each card say the family out loud: weak, or strong. It needs no table, no chips and no screen, and it belongs in the category of [practice away from a screen](https://21trainer.app/articles/blackjack-practice-away-from-screen/): a minute in a waiting room is enough. When the family call is instant, extend it: after weak or strong, say what the stiff hands do against that card.

**The upcard-only flash.** Hold the upcard fixed and interrogate it. Name its family, its bust-rate neighborhood, what hard 16 does against it, what soft 18 does against it. Ten cards, a few seconds each. It isolates knowledge a full-hand drill, where the upcard changes every deal, only touches in passing, and it pairs naturally with our [soft hand recognition drill](https://21trainer.app/articles/soft-hand-recognition-drill/), which isolates the other input the same way.

**One-row sweeps.** Take a single hand and recite its play against all ten upcards, in order, then check each answer against the chart. Soft 18 is the classic row: in this chart it doubles if allowed, otherwise stands, against 3 through 6, stands against 2, 7 and 8, and hits against 9, 10 and ace, which our piece on [soft 18](https://21trainer.app/articles/soft-eighteen-blackjack/) unpacks. Hard 12 is another good row, and 9,9 a third. The sweep teaches the row as a shape across the seam.

**Neighbor contrasts.** Drill the pairs of columns that sit across a boundary: 6 versus 7, the loudest step in the chart, and 8 versus 9, where only hard 16 and soft 18 move. Hold a fixed hand, turn up one of the two upcards at random, and answer at once; the gap this fills is the one our article on [blocked versus mixed drilling](https://21trainer.app/articles/mixed-hand-drills/) describes from the hand side.

## How you know it has landed

The test is not accuracy on a quiz of upcards, which tops out quickly; it is what happens to whole-hand decisions once the read is free. Our piece on [decision speed](https://21trainer.app/articles/blackjack-decision-speed/) sets the practice goal: a decision that feels like recognition rather than calculation, within a second or two of the cards landing. When the family read costs nothing, that budget goes entirely to the hand, and borderline rows like soft 18 stop eating it.

The second sign is structural rather than fast. Our article on [knowing when the chart is finally learned](https://21trainer.app/articles/when-is-the-chart-learned/) describes accuracy that holds across mixed order, with no stubborn cells and a rule set you can name. The upcard read is part of what makes that possible: some chart errors are not forgotten answers but answers recalled against the wrong family.

## Frequently asked questions

### Which dealer upcards are weak in blackjack?

The 2 through 6. Per the six deck stands on soft 17 tables, a 6 up busts about 42.3 percent of the time and a 5 about 41.8, while the low cards' bust rates all sit in the mid 30s or above. The 7 drops to about 26.2 percent, which is why the chart treats 2 through 6 as one group, weak, and 7 through ace as the other, strong. Fluent players read the family before anything else.

### Why does basic strategy change so much between a 6 and a 7?

Because the dealer's bust rate falls off a cliff there. Against a 6 the dealer busts about 42.3 percent of the time; against a 7 about 26.2, since a 7 reaches a standing total with one ten value card. Counting rows in the four to eight deck chart where the dealer stands on soft 17, 15 of the 28 rows play differently against a 7 than against a 6, more than half the chart.

### How do you practice reading dealer upcards?

Call the family before the play: flip a card, say weak or strong out loud, then say what the stiff hands do against it. Follow with one row sweeps, where you take a single hand such as soft 18 and recite its play against all ten upcards, and neighbor contrasts such as 6 versus 7. In the 21 Trainer app every practice hand gives instant feedback with an explanation, and heat maps by hand type and dealer upcard are in the Pro upgrade.

### Do any dealer upcards play identically?

Yes. In the four to eight deck chart with the dealer standing on soft 17, the 5 column and the 6 column are identical: not one of the 28 rows changes between them. That is the only neighboring pair with zero differences in that chart. It does not hold everywhere, so treat it as a fact about that chart rather than about the game, and always practice against the chart for your rules.

### How fast should you read a dealer upcard?

Instantly, the way you read a stop sign. The upcard read is the first half of every decision, and if it takes effort, the decision that follows cannot be automatic. A practice goal from our piece on decision speed: the full decision should feel like recognition rather than calculation within a second or two. If naming the family behind a given upcard ever takes thought, drill the reads on their own before drilling whole hands.

## Keep reading

- [Reading the Dealer Upcard](https://21trainer.app/articles/dealer-upcard-guide/)
- [Playing Against a Dealer 7, Hand by Hand](https://21trainer.app/articles/basic-strategy-vs-dealer-seven/)
- [Spotting Soft Hands Instantly: A Recognition Drill](https://21trainer.app/articles/soft-hand-recognition-drill/)