The idea in one sentence
Conditional probability is the chance of something happening given that you already know something else. In blackjack the “something else” is nearly always a set of cards you can see. Each visible card removes one possibility from the pool, so the chance of the next card has to be recalculated from what remains.
The working method is the same each time: count the unseen cards, count how many of them satisfy the question, and divide. All puzzles below use a single 52-card deck for simplicity, with 16 ten-valued cards and 4 aces.
Puzzle 1: a ten after a ten
Before any cards are dealt, the chance that the top card is ten-valued is 16 / 52, about 30.77%. Now suppose the first card you see is a ten-valued card. What is the chance that the next card is also ten-valued?
Fifty-one cards remain and 15 of them are ten-valued.
15 / 51 = 29.41%
The chance dropped slightly, from 30.77% to 29.41%, because one ten left the deck. As a quick follow-up, the chance that the first two cards are both ten-valued is the product: 16 / 52 x 15 / 51 = 240 / 2,652, about 9.05%. This is the chain rule of probability: the chance of A and then B is the chance of A multiplied by the chance of B given A.
Puzzle 2: using all the visible cards
You hold a ten and a six, and the dealer shows a five. You have seen three cards, so 49 are unseen. What is the chance that the next card is ten-valued?
Ten-valued cards: 16 - 1 (your ten) = 15.
15 / 49 = 30.61%
Next, what is the chance that hitting would bust you? A total of 16 busts on any card worth 6 or more. Count the unseen ones:
- Sixes: 4 - 1 (yours) = 3
- Sevens, eights and nines: 4 each = 12
- Ten-valued cards: 15
3 + 12 + 15 = 30 cards, so 30 / 49 = 61.22%. (The dealer’s five is not in this group, so it does not need adjusting.) The bust probability for a particular hit is therefore a conditional probability that depends on the exact hand, and it explains why the plays around hard sixteen are so awkward.
Puzzle 3: at least one ace
You are told that a two-card hand contains at least one ace. What is the chance it is a natural?
Count the hands. The number of two-card hands with no ace is 48 x 47 / 2 = 1,128 out of 1,326, so hands with at least one ace number 1,326 - 1,128 = 198. As a check, 4 x 48 = 192 hands have exactly one ace and 6 have two, which also gives 198.
Of those 198 hands, the naturals are the 64 that pair an ace with a ten-valued card (4 x 16).
64 / 198 = 32.32%
Compare that with the unconditional 64 / 1,326 = 4.83%. Learning that an ace is present raises the chance of a natural sharply, because you have ruled out most of the hands in which it could not happen. This is the reasoning behind many “given that” questions: restrict the pool to what is consistent with the information, then count.
Puzzle 4: what a dealer peek tells you
The dealer shows a ten-valued card and checks the hidden card for a natural. The check says there is no natural. What is the chance that the hidden card is ten-valued?
The hidden card is one of the 51 other cards in the deck, and the peek has ruled out the 4 aces, leaving 47 candidates. Of those, 15 are ten-valued (16 minus the upcard).
15 / 47 = 31.91%
Before the peek, that same card had a 15 / 51 = 29.41% chance of being ten-valued. Learning “not an ace” nudged it upward slightly. The peek also shows why no hole card games differ: without that check, you act before this information exists. The probability of the dealer’s natural itself is the subject of the dealer natural article.
A common trap: confusing sampling methods
All of these puzzles remove cards without putting them back, which is how a real deck behaves between shuffles. That differs from drawing with replacement, as with dice or coin flips, where earlier results do not affect later ones. Mixing the two gives wrong answers. With a coin, a run of heads does not make tails more likely, and believing it does is the gamblers fallacy. With a partly dealt deck, earlier cards genuinely change what is left, but the connection works through the cards that are gone, not through any sense of luck “evening out”.
Counting systems exploit that second fact in a limited way, tracking the removed cards to estimate the deck’s makeup, a topic covered under card removal. Even then, the casino can reshuffle, rules vary, and results stay uncertain.
Practical takeaways
- Always say what you know. The same question has different answers depending on the cards visible.
- Count unseen cards, not the full deck, and subtract each visible card from the right category.
- Check your answer by trying a second route, such as the complementary count in Puzzle 3.
Blackjack keeps a house edge regardless of how well you can calculate. These puzzles help explain the game; they do not guarantee any result. Gambling can be harmful, and anyone struggling can contact a local support service.