How Dealer Rules Affect Bust Probabilities
Dealer behavior is not arbitrary: it follows strict house rules that determine whether the dealer hits or stands on certain totals (most commonly “hit until 17” and either “stand on all 17s” or “hit on soft 17”). Those rules fundamentally shape bust probabilities. For example, if the dealer must hit soft 17 (an Ace counted as 11 plus other cards totaling six), the dealer will take one more card in situations where they would otherwise stand under the “stand on soft 17” rule; this increases the dealer’s expected final total and typically raises the dealer’s overall bust rate slightly. Conversely, any rule that forces the dealer to stand earlier reduces the chance of busting.
The number of decks in play also matters. In single-deck games card removal has a larger impact on remaining card composition, whereas in multi-deck shoes probabilities converge toward the infinite-deck, fixed-proportion expectation. In practice, bust probabilities against a given upcard shift a little when moving from one deck to six or eight decks. Penetration and shuffle frequency change how helpful card-counting adjustments are, but the baseline dealer bust probabilities by upcard remain qualitatively similar across deck counts. Other variants—such as rules that allow late surrender, dealer peeking for blackjack, or payout differences for blackjack—do not directly change the dealer’s bust probability but do affect the overall expected value of different player actions, which indirectly changes strategy choices informed by bust odds.
Dealer Upcard and Bust Odds: Percentage Estimates by Upcard
Understanding typical dealer bust probabilities by upcard gives players a quick heuristic for making decisions. The widely cited approximate bust probabilities for a dealer standing on soft 17 (using standard multi-deck odds) are: dealer Ace ~11.6%, 10 upcard ~21.4%, 9 ~23.3%, 8 ~23.9%, 7 ~25.9%, 6 ~42.1%, 5 ~42.9%, 4 ~40.2%, 3 ~37.6%, and 2 ~35.3%. Note these numbers are approximate and vary modestly by deck count and S17 vs H17 rules, but the pattern is robust: the dealer is most likely to bust when showing a 5 or 6, and least likely to bust when showing a 10 or Ace.
Why this pattern emerges: low upcards like 2–6 force the dealer to take more hits to reach a standing total of 17 or higher, increasing opportunities to exceed 21. The “bustiest” upcards (5 and 6) are especially weak because the dealer’s hole card often leaves them with a stiff total (12–16) requiring a hit, and many drawn cards will push them over 21. High upcards (10 or Ace) already carry strong value: a 10 upcard often means the dealer already has 20 with a face-down 10 or will easily reach 20–21; an Ace allows a flexible soft total where the dealer can often draw to a safe standing value without busting. These probabilities help explain why basic strategy often recommends conservative play (standing on marginal totals) against 5 and 6 and more aggressive plays (hitting or doubling in different spots) against dealer 7–Ace.
Using Bust Probabilities to Inform Player Strategy
Bust probabilities are a cornerstone of basic strategy reasoning. The core idea: when the dealer has a high probability of busting, the player benefits from minimizing their own risk of busting and letting the dealer self-destruct. Concretely, when the dealer shows a 5 or 6, basic strategy frequently advises standing on lower totals (for example, standing on 12–16 depending on rules and whether doubling or surrender options exist) rather than hitting, because the dealer is vulnerable to busting. Conversely, when the dealer shows a 7 or higher, the dealer is less likely to bust and more likely to make a strong final hand, so the player often needs to be more aggressive to improve their own total (e.g., hitting a 12–16 against a dealer 7).
Bust probabilities also influence doubling and splitting strategy. For instance, doubling down against dealer 5 or 6 is commonly profitable on hands like 9–11 because the high dealer bust chance improves the expected value of increasing the wager. Splitting tens is generally poor even vs weaker upcards because your two new hands often start worse than the original 20, and the dealer’s bust chance doesn’t justify weakening a very strong starting total. Insurance is another decision tied to dealer-upcard odds: the insurance bet is based on the probability the dealer has a ten-valued hole card when showing an Ace; despite the dealer’s Ace looking dangerous, insurance has negative expected value for most players because the insurance payoff (2:1) does not compensate for the true probability of a dealer blackjack in fair shuffled decks.
Card counting refines strategy by tracking composition: removing ten-value cards reduces dealer blackjack and the likelihood of dealer strong totals, which changes the conditional bust odds and can make some deviations from basic strategy + bet sizing profitable. But for most players, the simple upcard-based bust odds provide a practical, easy-to-apply guide for optimal play in typical casino games.
Exact vs. Approximate Methods: Monte Carlo, Combinatorics, and Recursive Calculations
There are three common ways to compute dealer bust probabilities: exact combinatorial/recursive calculations, Monte Carlo simulation, and approximate heuristics. Exact methods use dynamic programming or recursion to account for all possible hole cards and subsequent draws given a specific deck composition. They produce precise probabilities for a given rule set (number of decks, S17 vs H17) by enumerating the branches of the dealer’s hitting tree and aggregating probabilities. These methods are exact but can become computationally intensive for large shoe sizes unless implemented efficiently (memoization and state compression help).
Monte Carlo simulation runs many random-deal trials following dealer rules, then tallies bust outcomes. Simulations are flexible and easy to code; they can incorporate complex rules or nonstandard shoe compositions (e.g., after a few hands of play) and give statistically reliable estimates with enough runs. The downside is that simulations provide approximate results with a margin of sampling error; obtaining high precision requires many iterations.
Approximate heuristics rely on precomputed tables and analytic approximations (like the percentage-by-upcard table mentioned earlier). These are what most players and strategy charts use because they are simple and stable across common rule sets. They are sufficient for basic strategy and decision-making at the table. Advanced players or researchers who want exact numbers under precise shoe conditions use recursive calculations or targeted Monte Carlo runs. In practice, the method chosen depends on the level of precision needed: for casual play, the approximate upcard percentages and standard basic strategy are enough; for professional play, exact computation and card-counting-driven deviations become valuable.
