The question hides a more useful one
When someone asks how many possible combinations a dealer has in blackjack, they are usually standing at a table, looking at the dealer's upcard, and trying to work out what could happen next. The raw count of combinations is enormous, but it is not what decides the hand. What matters is how those combinations are distributed, because the dealer does not choose among them freely. The dealer follows a fixed drawing rule, and that rule collapses thousands of possible sequences into a small set of final outcomes.
A standard deck is 52 cards. If you want to count every possible sequence of cards the dealer could draw from a fresh shoe, the number runs into the millions once you account for different orders. But blackjack is not played by counting sequences. It is played by comparing final totals, and the dealer's final total is constrained by one mechanical rule: draw to 16, stand on 17, with the soft-17 variation depending on the table. That rule is what turns a huge combinatorial space into a small, knowable set of results.
Starting hands are the narrow part
The dealer's first two cards are the simplest thing to count, and the only part of the dealer's hand you can see before acting. One card is dealt face up, one face down. The upcard can be any of 13 ranks: ace through king. The hole card can be any of the remaining 51 cards, but for practical purposes what matters is the rank, not the specific card. A ten-valued card is any ten, jack, queen or king, and there are 16 of those in a single deck.
So the dealer's starting hand, expressed in ranks, is one of 13 upcards paired with one of 13 possible hole-card ranks. That gives 169 rank-pairs, though not all are equally likely. A ten-valued hole card is four times as common as a seven, because four ranks count as ten. This weighting is the first place the raw combination count misleads: 169 rank-pairs does not mean each pair happens with equal probability.
| Upcard | Hole card possibilities | What changes |
|---|---|---|
| Ace | Any rank, including another ace | Dealer may have a natural; insurance decision |
| Ten-value | Any rank | High chance of a strong starting total |
| 7 through 9 | Any rank | Dealer likely to stand after one or two draws |
| 2 through 6 | Any rank | Dealer must draw; bust risk rises |
The number that shapes your decision is not the total count of possible hole cards, but the proportion of them that produce a dealer bust, a made hand, or a forced draw. A dealer showing a six is dangerous to you not because there are many combinations, but because the combinations that beat you are outweighed by the combinations where the dealer draws to a stiff total and breaks.
Final totals collapse into a handful of results
Once the dealer starts drawing, the number of possible card sequences grows quickly, but the number of final outcomes does not. The dealer can finish with a total of 17, 18, 19, 20, 21, or a bust over 21. That is six outcome classes. A dealer cannot finish on 16 or lower, because the drawing rule forces another card. A dealer cannot finish on 22 as a playable total, because that is a bust.
This is the key structural fact: thousands of card sequences map onto six possible endings. The interesting number is not how many sequences exist, but what fraction of them land in each ending, given the upcard you can see. A dealer showing a ten has a much higher share of sequences ending in 20 or 21 than a dealer showing a four. A dealer showing a five or six has the highest share of sequences ending in a bust. These shares are probabilities, not counts, and they are what basic strategy responds to.
Why the count itself does not help you win
Counting combinations is a descriptive exercise. It tells you what can happen, not what tends to happen, and certainly not what will happen on this hand. The house edge in blackjack is not built from the number of possible dealer hands. It is built from the rules: the player acts first and can bust before the dealer ever draws, the dealer's fixed drawing rule removes choice, and a natural blackjack pays less than even money at most tables. Those structural features produce the edge, not the size of the combinatorial space.
You can see this by comparing two games with very different combination counts. A slot machine has millions of possible reel positions, far more than a blackjack shoe. A single-number roulette bet has 37 or 38 possible outcomes. The number of combinations does not set the edge. The payout ratio against the probability of winning sets the edge. In blackjack, the dealer's constrained drawing rule and the player's first-mover disadvantage set it.
Where the arithmetic lives
Suppose, as an example, you want the probability that a dealer showing a six will bust. You would enumerate every sequence of cards the dealer could draw from the remaining shoe, apply the stand-on-17 rule, and count the fraction that end over 21. The count depends on the exact shoe composition at that moment, which is why published dealer-bust tables are approximations for a full shoe and shift as cards are removed. The table below shows the shape of the relationship, not fixed percentages, because the exact figures depend on deck count and composition.
| Dealer upcard | Bust tendency | Player response logic |
|---|---|---|
| 2 through 6 | Higher bust share | Stand on weaker totals, let the dealer draw |
| 7 through 9 | Moderate bust share | Hit stiff totals, stand on made hands |
| Ten-value or ace | Lower bust share | Take more risk to build a competitive total |
The dealer's upcard is a signal about the distribution of final outcomes, not a guarantee. A six does not mean the dealer will bust. It means the share of sequences ending in a bust is larger than it is against a ten. That is the entire informational content of the upcard, and it is enough to build a basic strategy around.
Short sessions do not look like the distribution
Even when you know the distribution of dealer outcomes, a single session will not reproduce it. If the dealer busts against a six on three consecutive hands, that is not the long-run share asserting itself. It is variance. A distribution describes what happens over a very large number of independent hands, not what happens in the next twenty minutes. This is the same reason RTP and house edge are long-run statistical averages rather than per-session guarantees.
The dealer's hand is one half of a two-party comparison. Your own decisions change which of your possible totals you bring to the comparison, and they change how often you bust before the dealer acts. What they do not change is the dealer's drawing rule or the distribution of dealer outcomes given the upcard. That is the boundary of player control, and it is worth being precise about it.
What the question can and cannot answer
No single article can give you the exact number of dealer combinations for your specific table, because the count depends on the number of decks in the shoe, the cards already dealt, and whether the dealer stands or hits on soft 17. What a structural answer can give you is the framework: 13 possible upcards, six possible final outcome classes, and a drawing rule that makes the distribution of those outcomes knowable in principle and uneven in practice.
If you want the precise bust probability for a given upcard under a given rule set, the source is a published basic-strategy or dealer-outcome table matched to your table's rules: deck count, soft-17 rule, and whether surrender or other options change the comparison. The game information panel or the table's posted rules will state the deck count and soft-17 rule. That is the document that carries the numbers for your specific game.
FAQ
How many different starting hands can a blackjack dealer have?
In rank terms, 169 upcard-and-hole-card pairs, from ace-ace through king-king. The pairs are not equally likely, because ten-valued cards are four ranks in one, so a ten-valued hole card appears far more often than any single other rank.
Does the dealer have more possible hands than the player?
The raw combinatorial space is the same size for both, since both draw from the same shoe. What differs is the rule set: the player chooses when to draw and may double or split, while the dealer must draw to 16 and stand on 17. That fixed rule concentrates the dealer's outcomes into a smaller set of final totals.
Why do dealer bust probabilities matter more than the combination count?
Because the count alone tells you nothing about how often each outcome occurs. The bust probability attached to each upcard is what drives basic strategy decisions, and that probability is a share of sequences, not a total number of them. A large combinatorial space with a skewed distribution behaves very differently from a uniform one.
Play responsibly
Gambling should be treated as paid entertainment, never as a way to earn income or recover losses. 1-800-MY-RESET (1-800-697-3738)
18+ or 21+ depending on where you are; follow the minimum age that applies to you.
This article is general information about how these mechanics work. It is not legal advice and not a recommendation to gamble or to use any particular operator. Availability and legality differ by jurisdiction — check the rules that apply where you are.