Further reading — Module 06

Perfect play in blackjack

Blackjack is unusual among casino games because the player makes decisions that change the mathematics of the hand. "Perfect play" is the shorthand for making the statistically correct decision every single time — and it is the only reason the house edge on blackjack is measured in fractions of a percent rather than whole digits.

What perfect play actually means

Perfect play — usually called basic strategy — is the mathematically optimal action (hit, stand, double, split, or surrender) for every possible combination of the player's hand and the dealer's up-card. It is not intuition, not card counting, and not a system for winning. It is the single decision, in each of roughly 300 possible situations, that produces the highest expected value against the fixed rules of the game.

Basic strategy is derived by simulating every possible outcome for each decision and picking the action whose average result loses the least (or wins the most). Because the dealer's behaviour is fixed by house rules — typically "hit on 16, stand on 17" — the maths resolves to one correct answer per situation. There is no ambiguity, and no room for feel.

How perfect play shapes the house edge

Under common multi-deck rules with basic strategy played flawlessly, the house edge on blackjack sits around 0.5%. That is by far the smallest edge of any mainstream table game. But that number assumes every decision is correct — every hit, every stand, every double, every split.

The moment a player deviates, the edge widens. Studies of casual play typically measure the house edge at somewhere between 1.5% and 2.5% — three to five times the theoretical minimum. The deviations are rarely dramatic; they are small, familiar mistakes repeated across thousands of hands:

• Standing on a 16 against a dealer 7 or higher.
• Refusing to split 8s or aces.
• Taking insurance on a strong hand.
• Doubling only when the hand "feels" strong.
• Hitting a soft 18 against a dealer 9 or 10.

Each mistake is worth a fraction of a percent on its own. Together, over a session, they compound into the same expected loss as playing a much worse game correctly.

Why rules matter as much as strategy

Perfect play only reaches its ~0.5% edge under favourable rules. The same flawless decisions produce very different expected values depending on the table:

• Blackjack paying 3:2 keeps the edge near 0.5%. A table paying 6:5 for a natural blackjack raises the edge by roughly 1.4% — larger than every other rule change combined.
• Dealer stands on soft 17 is better for the player than dealer hits soft 17 by about 0.2%.
• Double after split and late surrender each shave a small amount off the edge.
• Fewer decks are marginally better for the player than more decks, all else equal.

Perfect play does not overcome unfavourable rules. It only extracts the smallest possible edge from whichever rule set is on offer.

The limit of perfect play

Even with textbook decisions on a favourable table, the expected value of every hand is still negative. Perfect play does not turn blackjack into a positive-EV game — it simply produces the least negative version of it. Over a long enough sample, a player using basic strategy will still lose roughly £0.50 for every £100 wagered. The variance around that number is wide in the short term, which is what allows winning sessions to exist, but the long run is fixed.