iGaming Journalist & Crypto Casino Analyst
Blackjack basic strategy is built almost entirely on one fact: the dealer's visible card determines how likely the dealer is to bust. This page publishes the exact probability of every dealer outcome (17, 18, 19, 20, 21 and bust) for every upcard, for one to eight decks, under both stand-on-soft-17 (S17) and hit-on-soft-17 (H17) rules. The figures come from an exact recursive calculation rather than a simulation, so there is no sampling error. Overall, the dealer busts 28.19% of the time in a six-deck S17 game, but that single number hides a huge range, from 16.7% when the upcard is an ace to 42.3% when it is a six.
Key Findings
The dealer busts 28.19% of hands in a six-deck S17 game and 28.58% under H17. A six upcard busts 42.3% of the time (peeked), a ten only 23.0%. Dealer blackjack occurs in 4.75% of hands, and the number of decks changes the overall bust rate by less than 0.2 percentage points.
Why Dealer Outcome Probabilities Matter
Unlike the player, the dealer has no choices. House rules force the dealer to draw until reaching at least 17, and, depending on the casino, to either stand or hit on a soft 17 (a hand containing an ace counted as 11, such as ace-six). That fixed procedure turns the dealer's hand into a pure probability problem: given one visible card, what is the distribution of final totals? Every decision in basic strategy, from doubling down on 11 against a six to standing on 12 against a four, is an application of that distribution. If you have ever wondered why the chart tells you to stand on a stiff hand against a weak dealer card, the answer sits in the bust column of the tables below.
It is also a useful reference for writers, students and analysts. The numbers here are commonly quoted in rounded form (for example "the dealer busts about 28% of the time"), but a precise breakdown by upcard, deck count and soft-17 rule is harder to find in one citable place. For a practical guide to applying these probabilities at the table, see our online casino guides, and for a broader introduction visit the DeucesCracked casino hub.
Methodology
Exact calculation, not simulation
We computed every probability with an exhaustive recursion over the composition of the remaining shoe. Starting from a freshly shuffled shoe of one, two, four, six or eight 52-card decks, we remove the dealer's upcard and then enumerate every possible sequence of draws according to the dealer's rule, weighting each draw by the exact count of cards remaining at that point. This means cards are drawn without replacement, so the calculation correctly captures the "composition-dependent" effect that is why single-deck and eight-deck figures differ slightly. The results are exact up to floating-point precision, not estimates.
We report two kinds of figure. Unconditional (or "no-peek") probabilities include the chance that the dealer has a natural blackjack and are used for overall dealer outcomes and for the ace and ten upcards in the blackjack table. Peeked probabilities exclude dealer blackjack and apply when the dealer checks the hole card for a natural and the hand continues only if no natural exists, which is standard in American-style games. Peeked figures are the ones relevant to a player who has already seen the dealer fail to show a blackjack.
Assumptions
The calculation assumes a fresh shoe with only the dealer's upcard removed. It does not remove the player's cards or other players' cards, which would modestly change the probabilities for any individual hand; it also assumes no cut-card effect, no card counting information and no continuous shuffling adjustments. Rules are the standard dealer procedure: draw on 16 or less, stand on hard 17 or more, and either stand (S17) or hit (H17) on soft 17. As a cross-check, our overall dealer blackjack rate of 4.75% is consistent with the "approximately 4.8% of hands" figure given in the Wikipedia blackjack article, and our single-deck S17 overall bust rate of 28.36% matches the commonly quoted figure of roughly 28.4%.
Dealer Final Totals by Upcard: Six Decks, Stand on Soft 17
This is the core reference table. Each row shows the probability of each final dealer outcome given the visible upcard, assuming six decks, S17 and a peeked hole card (no dealer blackjack). Rows sum to 100%.
| Upcard | 17 | 18 | 19 | 20 | 21 | Bust |
|---|---|---|---|---|---|---|
| A | 18.8% | 18.9% | 18.9% | 18.9% | 7.7% | 16.7% |
| 2 | 14.0% | 13.4% | 13.0% | 12.4% | 11.8% | 35.4% |
| 3 | 13.4% | 13.1% | 12.5% | 12.1% | 11.5% | 37.4% |
| 4 | 13.1% | 12.4% | 12.1% | 11.6% | 11.2% | 39.6% |
| 5 | 12.2% | 12.2% | 11.8% | 11.2% | 10.8% | 41.8% |
| 6 | 16.6% | 10.6% | 10.6% | 10.2% | 9.7% | 42.3% |
| 7 | 36.9% | 13.8% | 7.8% | 7.9% | 7.4% | 26.2% |
| 8 | 12.9% | 36.0% | 12.9% | 6.9% | 6.9% | 24.4% |
| 9 | 12.0% | 11.7% | 35.2% | 12.0% | 6.1% | 22.9% |
| 10/J/Q/K | 12.1% | 12.1% | 12.1% | 36.8% | 3.8% | 23.0% |
How to read it
Two groups of upcards emerge immediately. The "bust cards" are 2 through 6, where the dealer busts between 35.4% and 42.3% of the time, and the "strong cards" are 7 through ace, where the dealer is far less likely to bust. The probability of busting rises steadily from a two to a six, then falls sharply at a seven, because a seven is already a pat-17 hand in the dealer's favour when the hole card is a ten. Notice how the distribution for 7, 8 and 9 has a single dominant peak (36.9% 17 for a seven, 36.0% 18 for an eight, 35.2% 19 for a nine), because a ten-value hole card (which accounts for 16 of every 52 cards) produces that exact total at once. The ten upcard peaks at 20 (36.8%), reflecting the same mechanism.
The ace is unusual. Once dealer blackjack is excluded, an ace upcard has the most evenly spread distribution of any card: 18.8% to 18.9% for each of 17, 18, 19 and 20, and a bust probability of only 16.7%, because a soft hand can absorb any hit without busting. Yet the ace is the hardest card for the player when blackjack is included, since the dealer holds a natural in 30.9% of ace-up hands before peeking.
Dealer Bust Probability by Upcard and Number of Decks
Deck count matters less than most players think. The table shows bust probability (peeked, S17) for each upcard as the shoe grows from one deck to eight, followed by the six-deck and eight-deck H17 figures.
| Upcard | 1 deck S17 | 2 decks S17 | 4 decks S17 | 6 decks S17 | 8 decks S17 | 6 decks H17 | 8 decks H17 |
|---|---|---|---|---|---|---|---|
| A | 16.98% | 16.81% | 16.73% | 16.70% | 16.69% | 20.13% | 20.11% |
| 2 | 35.30% | 35.33% | 35.35% | 35.35% | 35.35% | 35.67% | 35.67% |
| 3 | 37.56% | 37.48% | 37.43% | 37.42% | 37.41% | 37.70% | 37.69% |
| 4 | 40.28% | 39.85% | 39.65% | 39.58% | 39.55% | 39.85% | 39.81% |
| 5 | 42.89% | 42.25% | 41.94% | 41.84% | 41.79% | 41.96% | 41.91% |
| 6 | 42.08% | 42.21% | 42.27% | 42.28% | 42.29% | 43.93% | 43.93% |
| 7 | 25.99% | 26.11% | 26.17% | 26.19% | 26.20% | 26.19% | 26.20% |
| 8 | 23.86% | 24.16% | 24.32% | 24.37% | 24.40% | 24.37% | 24.40% |
| 9 | 23.34% | 23.09% | 22.97% | 22.92% | 22.90% | 22.92% | 22.90% |
| 10/J/Q/K | 23.25% | 23.11% | 23.05% | 23.02% | 23.01% | 23.02% | 23.01% |
Moving from one deck to eight changes the bust probability by at most about 1.1 points for any upcard (the largest shifts are for the 4 and 5), and by well under half a point for most others. The composition effect is strongest for middle cards: with a single deck, a five upcard busts 42.89% versus 41.79% with eight decks. For the six, the figures are 42.08% and 42.29%. For practical purposes, the six-deck and eight-deck columns are almost identical, which is why published basic-strategy charts rarely distinguish between them. The Wikipedia blackjack article reports a house advantage that rises from 0.16% with a single deck to 0.66% with eight decks under one specific ruleset, so deck count matters far more for the player's overall edge than for the dealer's bust rate: other rule interactions, such as the player's ability to double and split, drive the difference.
Hit vs. Stand on Soft 17: What Changes
Under H17 rules the dealer must hit a soft 17, such as ace-six, which gives the dealer another chance to improve a weak total but also another chance to bust. The Wikipedia article estimates that switching from H17 to S17 reduces the house edge by about 0.2% in the player's favour. The table below shows where those differences come from, for a six-deck shoe, peeked.
| Upcard | Bust S17 | Bust H17 | Change (pp) | Final 17 S17 | Final 17 H17 |
|---|---|---|---|---|---|
| A | 16.70% | 20.13% | +3.42 | 18.8% | 8.3% |
| 2 | 35.35% | 35.67% | +0.32 | 14.0% | 13.0% |
| 3 | 37.42% | 37.70% | +0.28 | 13.4% | 12.6% |
| 4 | 39.58% | 39.85% | +0.27 | 13.1% | 12.2% |
| 5 | 41.84% | 41.96% | +0.12 | 12.2% | 11.8% |
| 6 | 42.28% | 43.93% | +1.64 | 16.6% | 11.5% |
| 7 | 26.19% | 26.19% | +0.00 | 36.9% | 36.9% |
| 8 | 24.37% | 24.37% | +0.00 | 12.9% | 12.9% |
| 9 | 22.92% | 22.92% | +0.00 | 12.0% | 12.0% |
| 10/J/Q/K | 23.02% | 23.02% | +0.00 | 12.1% | 12.1% |
The rule matters most for the upcards that most often produce a soft 17: the ace and the six (a six plus a hole ace), though any upcard can reach a soft 17 after several draws. The ace upcard shows the biggest change: bust probability rises from 16.70% under S17 to 20.13% under H17, a gain of 3.42 points, while the frequency of a final 17 falls from 18.8% to 8.3%. A six upcard busts 43.93% under H17 versus 42.28% under S17. The dealer's extra draws convert some soft 17s into 18 through 21 (good for the house) and others into busts (good for the player), but on balance the dealer improves more often than they bust, which is why H17 favours the house. At the whole-table level, the H17 rule raises the overall bust rate by about 0.38 points, from 28.19% to 28.58%, yet the dealer also finishes with 20 or 21 more often, which is the larger effect.
Overall Dealer Outcome Distribution
Weighting each upcard by its frequency in the shoe (four cards of each rank from ace to nine and sixteen ten-value cards per deck, out of 52) produces the overall distribution of dealer outcomes across all hands. These are unconditional probabilities that include dealer blackjack as a separate category.
| Decks | Rule | 17 | 18 | 19 | 20 | 21 (not blackjack) | Blackjack | Bust |
|---|---|---|---|---|---|---|---|---|
| 1 | S17 | 14.58% | 13.81% | 13.48% | 17.58% | 7.36% | 4.83% | 28.36% |
| 1 | H17 | 13.44% | 13.97% | 13.68% | 17.78% | 7.56% | 4.83% | 28.75% |
| 2 | S17 | 14.55% | 13.88% | 13.41% | 17.81% | 7.32% | 4.78% | 28.26% |
| 2 | H17 | 13.39% | 14.06% | 13.61% | 18.00% | 7.51% | 4.78% | 28.64% |
| 4 | S17 | 14.53% | 13.91% | 13.38% | 17.92% | 7.29% | 4.76% | 28.21% |
| 4 | H17 | 13.36% | 14.11% | 13.58% | 18.12% | 7.49% | 4.76% | 28.59% |
| 6 | S17 | 14.52% | 13.93% | 13.37% | 17.95% | 7.29% | 4.75% | 28.19% |
| 6 | H17 | 13.35% | 14.12% | 13.57% | 18.15% | 7.49% | 4.75% | 28.58% |
| 8 | S17 | 14.52% | 13.93% | 13.36% | 17.97% | 7.28% | 4.75% | 28.18% |
| 8 | H17 | 13.34% | 14.13% | 13.56% | 18.17% | 7.48% | 4.75% | 28.57% |
The most commonly quoted headline, that the dealer busts roughly 28% of the time, holds in every configuration: the range runs from 28.18% to 28.75%. A dealer who does not bust finishes at 17 through 21, and the most common single outcome is a bust, followed by 20 (about 18.0% in six-deck S17) and then 17 (14.5%). The least common finish is a non-blackjack 21 at 7.3%, followed by blackjack at 4.75%. Together, 21 in any form occurs in 12.0% of hands.
Dealer Blackjack Probability by Upcard
The dealer's natural is only possible with an ace or a ten-value card showing. The probability is simply the chance that the hole card completes the pair, but because it is drawn without replacement, it varies slightly with the number of decks.
| Decks | Dealer blackjack, ace up | Dealer blackjack, ten up | All hands |
|---|---|---|---|
| 1 | 31.37% | 7.84% | 4.83% |
| 2 | 31.07% | 7.77% | 4.78% |
| 4 | 30.92% | 7.73% | 4.76% |
| 6 | 30.87% | 7.72% | 4.75% |
| 8 | 30.84% | 7.71% | 4.75% |
With one deck, the odds are exactly 16 in 51 (31.37%) for an ace showing, and 4 in 51 (7.84%) for a ten-value card, because there are sixteen ten-value cards and four aces among the 51 unseen cards. As the number of decks grows, the figures converge on 16/52 (30.77%) and 4/52 (7.69%). Note that the "all hands" column falls from 4.83% to 4.75%, the same pattern the Wikipedia blackjack article describes in saying that blackjack occurs in approximately 4.8% of hands. The Wikipedia article also notes that the dealer has an ace in the hole less than one-tenth of the time when a ten-value card is showing: our figure is 7.7%.
Ranking Upcards by Dealer Weakness
| Rank (weakest first) | Upcard | Dealer bust (no peek) | Dealer ends 20 or 21 (peeked) | Dealer blackjack |
|---|---|---|---|---|
| 1 | 6 | 42.3% | 19.9% | 0.0% |
| 2 | 5 | 41.8% | 22.0% | 0.0% |
| 3 | 4 | 39.6% | 22.8% | 0.0% |
| 4 | 3 | 37.4% | 23.6% | 0.0% |
| 5 | 2 | 35.4% | 24.2% | 0.0% |
| 6 | 7 | 26.2% | 15.2% | 0.0% |
| 7 | 8 | 24.4% | 13.9% | 0.0% |
| 8 | 9 | 22.9% | 18.1% | 0.0% |
| 9 | 10/J/Q/K | 21.2% | 40.6% | 7.7% |
| 10 | A | 11.5% | 26.7% | 30.9% |
The ranking confirms the standard advice. The six is the dealer's weakest card (42.3% bust), followed by the five, four, three and two. The seven through ten, plus the ace, are strong. Of these, the ten is strongest in terms of ending with a 20 or 21 (40.6% when peeked), while the ace is strongest when blackjack is included. Basic strategy tells players to stand on stiff hands of 13 through 16 against a dealer 2 through 6 (and on 12 against a 4, 5 or 6) for precisely this reason: the dealer is likely enough to bust that the player's risk of busting by drawing is not worth taking. Against a 7 through ace, the player must assume the dealer will reach 17 or higher and so should draw to improve a stiff hand.
What These Probabilities Mean for Players
Standing on stiff hands
A player holding hard 16 against a dealer six wins whenever the dealer busts (42.3% peeked) and also if the dealer finishes with a total below the player's total. With 16 the player can only win by dealer bust, which is why standing is generally correct: hitting risks busting the hand about 62% of the time (any card that is a 6 or more busts, 8 out of 13 ranks, or roughly 61.5% before accounting for removed cards), whereas standing wins 42.3% of the time. Against a ten upcard, standing on 16 wins only when the dealer busts (23.0%), which is why the hand is among the worst in the game.
Doubling down
Doubling down is most attractive when the dealer is weak, because a doubled bet captures the large share of hands in which the dealer busts. The probability tables explain why strategy charts allow doubling a soft 17 against a three through six, and why a hard 11 is doubled even against high cards in most games: the hand has a strong chance of making a high total, and the dealer's bust rate matters less. For the complete decision tables, see our guides section.
Insurance and the natural
When the dealer shows an ace, insurance pays 2 to 1 on a wager that the hole card is a ten-value card. The true chance of a ten in the hole is 30.87% in a six-deck game, which is less than the one-in-three that 2-to-1 pays require to break even (33.33%). The insurance bet therefore carries a negative expectation for a player who cannot track the composition of the shoe, even though it feels like a bargain. For a deeper look at that bet and at surrender, see our blog for recent casino strategy articles.
Limitations
These probabilities describe the dealer's hand only, not the player's expected value. A complete basic-strategy analysis must also account for the player's own cards, doubling and splitting rules, the payout for a natural (3:2 versus 6:5), surrender options and the cut-card position. The calculations assume a fresh shoe with only the upcard removed; in a live game the cards already dealt to the player and other players change the remaining composition slightly, which can move individual probabilities by a few tenths of a percentage point. Results also assume an honest shuffle and standard dealer rules. Games with unusual rules, such as dealers who hit on all 17s, "Spanish 21" decks with the tens removed, or continuous shuffling machines, are not covered. Online live-dealer games generally follow the standard rules used here, but players should always check the rule card for the specific table.
Frequently Asked Questions
How often does the dealer bust in blackjack?
Across all upcards, the dealer busts 28.19% of the time in a six-deck game where the dealer stands on soft 17, and 28.58% where the dealer hits soft 17. The figure varies from 28.18% to 28.75% across all deck counts and both rules in our calculation.
Which dealer upcard is the weakest?
The six. It busts 42.3% of the time (unconditional, six decks, S17), more than any other card, followed by the five (41.8%) and the four (39.6%).
Does the number of decks change the dealer's bust rate?
Only slightly. Overall, the dealer busts 28.36% of the time with one deck and 28.18% with eight decks under S17, a difference of 0.17 percentage points. Deck count affects the player's edge mainly through other rule interactions.
How much does hitting soft 17 help the casino?
The Wikipedia blackjack article estimates that switching from H17 to S17 reduces the house edge by roughly 0.2%. In our calculation, H17 raises the overall dealer bust rate slightly (from 28.19% to 28.58%) but lets the dealer improve a soft 17 into a 20 or 21 more often, which is the larger effect for the house.
How likely is a dealer blackjack?
About 4.75% of hands in a six-deck shoe. If the upcard is an ace, the chance is 30.9%; if it is a ten-value card, 7.7%.
Sources and Further Reading
- Wikipedia, Blackjack (basic strategy house edge, H17 vs. S17 effect, blackjack frequency, house advantage by number of decks).
- Wikipedia, Blackjack: Basic strategy (strategy background).
- DeucesCracked original exact calculation (November 2026): dealer outcome distributions by upcard for 1, 2, 4, 6 and 8 decks under S17 and H17 rules, computed by exhaustive recursion without replacement.
Cite This Article
If you use data from this article, please link back to https://www.deucescracked.com/blog/blackjack-dealer-bust-probability-by-upcard-data. Suggested citation: DeucesCracked Editorial, "Blackjack Dealer Bust Probability by Upcard," November 2026.
Related Articles
Hot and Cold Slots: Why Slot Machine Streaks Are a Myth
Are slots ever hot or cold? Learn how RNGs and RTP work, why streaks are random, and which online slot habits actually protect your bankroll. Read the facts.
Nov 5, 2026Casino Loss Limits: How to Set a Session Budget That Works
Set casino loss limits and a session budget that protects your bankroll. Learn stop-loss rules, win goals and deposit limits to play online casinos responsibly.
Nov 5, 2026European vs American Roulette: House Edge Compared 2026
European vs American roulette house edge compared: 2.70% vs 5.26%, plus La Partage, En Prison and why systems fail. Choose the better wheel.
Nov 4, 2026Jacks or Better Strategy: Hold Order and Paytable Guide
Jacks or Better strategy for full-pay 9/6 video poker: paytable check, hold priorities, max coins and common mistakes. Play smarter with this guide.
Nov 4, 2026Join the Conversation
Be respectful. No spam. Strategy discussion welcome.