Reading the Dealer's Upcard: the Half of the Hand Everyone Ignores
New players stare at their own cards. Good players stare at the dealer's. Your two cards are only half the hand — the other half is the single card the dealer shows, and it carries a precise, published probability of how the dealer's hand ends. Every row of the basic strategy chart is really an answer to one question: what is that upcard threatening to become?
Put the upcard to work →The upcard is a probability forecast
The dealer has no decisions. The rules force the play — hit every total through 16, stand on every 17 — so once the upcard is visible, the distribution of the dealer's final hand is pure arithmetic. A dealer showing a 6 will finish on 17, 18, 19, 20, 21 or a bust with fixed, knowable frequencies, and those frequencies are published to the decimal in standard references like Wizard of Odds Appendix 1. This is what makes blackjack a strategy game rather than a guessing game: you are not playing against a mind, you are playing against a distribution, and the upcard tells you which distribution. Basic strategy is nothing more than the precomputed best response to each one. Ignore the upcard and you are throwing away exactly half the information on the table — which is why two players holding the same 14 should routinely make opposite decisions depending on what the dealer shows.
Dealer bust rates by upcard
Here is the forecast, card by card. The dealer's bust probability is highest behind the middle cards: roughly 42% showing a 5 or 6, about 40% showing a 4, around 37% for a 3 and 35% for a 2. Then it falls off a cliff. Showing a 7 the dealer busts only about 26% of the time, an 8 about 24%, a 9 about 23%, and a ten-value card about 21%. The ace is the safest card in the game for the house — with the blackjack peek resolved, a dealer ace busts only around 17%, and it simultaneously threatens the strongest finishing totals. Why do 5 and 6 bust the most? Because they force the dealer through the stiff zone: a ten in the hole makes 15 or 16, which the rules force the dealer to hit into a 54% to 62% bust probability of their own. Meanwhile a 7 through ten in the hole behind a strong upcard makes an instant pat 17 to 20. The dealer suffers the exact same bust ladder you do — the upcards just determine how often the rules march them into it.
Two modes: dealer weak (2-6) and dealer strong (7-A)
Collapse those ten numbers into two regimes and you have the operating system of basic strategy. Against a weak upcard, 2 through 6, the dealer busts roughly 35% to 42% of the time, so your default posture is: take no unnecessary risk and get more money on the table. Stand on your stiffs (12 through 16, with the small exception of 12 against a 2 or 3), double aggressively on 9 through 11 and on many soft hands, and split more pairs. You are essentially betting that the dealer self-destructs, and at these frequencies that is a bet worth pressing. Against a strong upcard, 7 through ace, the dealer finishes 17-or-better roughly three times in four, so passivity is fatal. Now you hit your stiffs despite the bust risk, double only on your premium totals, split conservatively, and surrender the true disasters where the option exists. Same cards in your hand, opposite plays — because the thing you are playing against changed.
Worked examples: 13 versus 6, 16 versus 10
Two contrasts make the logic concrete. Hold hard 13 against a dealer 6 and basic strategy says stand, even though 13 is a miserable total that wins nothing on its own. The reasoning: hitting 13 busts you about 39% of the time, while the dealer showing a 6 busts about 42% of the time without your help. Why take a 39% risk of instant death when the opponent will hand you the pot at 42%? Stay alive and let the rules do the work. Now hold hard 16 against a dealer ten. Standing means you win only via a dealer bust — about 21% — so standing loses close to four times in five. Hitting busts you 62% of the time, which is awful, but the 38% of draws that survive convert a near-dead hand into a live one often enough to lose less overall. Neither play feels good; one is simply less bad, and the difference is entirely the upcard. Where surrender is available, 16 against a ten is the textbook case for taking half your bet back — which is exactly why Llamabet offers it.
The doubling windows the upcard opens
The upcard does not just decide defense — it defines when to attack. Doubling down is profitable when your chance of finishing strong collides with a dealer likely to finish weak, and the weak upcards throw those windows wide open. Hard 11 doubles against nearly everything, hard 10 against dealer 2 through 9, and hard 9 against the soft middle of 3 through 6. Against a 5 or 6 — the twin peaks of dealer bust probability — the window expands further to many soft hands: A-2 through A-7 all double against a 6 in standard multi-deck strategy, plays that would be unthinkable against a strong card. The pattern generalizes: as the upcard weakens, the set of hands worth doubling grows, because every extra point of dealer bust probability subsidizes your risk. At a table like Llamabet, where the action set is hit, stand, double and surrender, the doubling windows are where the aggressive half of your edge management lives — the upcard tells you exactly when to open them and when to keep the extra bet in your wallet.
The 'ten in the hole' heuristic — useful, but know its limits
Old blackjack wisdom says: assume the dealer has a ten in the hole. As a memory aid it is genuinely useful — it instantly explains why a 6 is weak (assume 16, dealer must hit into a 62% bust) and a ten is strong (assume 20, beat it or lose). But treat it as a metaphor, not math. Ten-value cards are only 4 of 13 ranks, about 31% of the deck, so the dealer does not have a ten under there most of the time — and strategy computed against a literal guaranteed ten would misplay plenty of edges, like standing on stiffs against a 2 where the true bust rate is a healthy 35%. The correct mental model is the full distribution, which is exactly what basic strategy already encodes. Use the heuristic to remember the chart, never to override it. And at Llamabet you can verify the cards were never anyone's guess: every hand is dealt from a fresh four-deck shoe whose shuffle hash is committed on Sui before you bet and revealed after, dealer hits 16 and stands on 17, blackjack pays 3:2 from a 1 SUI minimum, and you can sign in with Google zkLogin or a Sui wallet like Slush with instant on-chain payouts.
Frequently asked questions
What dealer upcard busts the most in blackjack?
The 5 and 6, at roughly 42% each, with the 4 close behind around 40%. These cards force the dealer through the stiff zone — a ten in the hole makes 15 or 16, which the rules compel the dealer to hit at high bust risk. The safest upcards for the house are the ten (about 21% bust) and the ace (around 17% once the blackjack peek is resolved).
Why do you hit 16 against a dealer 10 but stand on 13 against a 6?
Because the comparison is always your bust risk versus the dealer's. Against a 6 the dealer busts about 42% of the time, so risking a 39% one-card bust on your 13 is unnecessary — stand and let them fail. Against a ten the dealer busts only about 21% of the time, so a standing 16 is nearly dead; hitting busts 62% of the time but loses less overall.
Should you assume the dealer has a 10 in the hole?
Only as a memory aid. Ten-value cards are about 31% of the deck, so the assumption is literally false most of the time, and strategy built on a guaranteed ten misplays several situations. Basic strategy already averages over the true distribution of hole cards — use the heuristic to recall the chart, not to second-guess it.