Blackjack: Basic Strategy, the House Edge and Why Card Counting Works

Basic strategy: a lookup table, not a hunch

Basic strategy is the mathematically optimal action — hit, stand, double down or split — for every combination of player hand and dealer up-card, derived by exhaustively computing the expected value of each choice against the probability distribution of a randomly shuffled shoe. It was first worked out rigorously by Baldwin, Cantey, Maisel and McDermott in 1956 using desk calculators, and refined by Edward Thorp's 1962 computer simulations in Beat the Dealer. Played perfectly, basic strategy alone brings the house edge down to roughly 0.5% depending on the exact table rules (number of decks, whether the dealer hits or stands on soft 17, doubling restrictions) — among the lowest edges of any game on a casino floor.

The table encodes non-obvious results: standing on a hard 12 against a dealer 4, 5 or 6 (because the dealer is likely to bust) but hitting the identical hard 12 against a dealer 7 through ace; doubling down on 11 against almost anything; splitting aces and eights nearly unconditionally. None of this is intuition — it falls straight out of the expected-value arithmetic over every possible next card and every possible dealer outcome.

Why the house still has an edge with perfect play

The dealer's structural advantage is simple and survives even perfect basic strategy: the player acts first. If the player busts, the house wins immediately regardless of what the dealer's hand would eventually have done — the dealer never has to reveal or resolve a hand that already lost. This ordering asymmetry is worth roughly 8% on its own; basic strategy claws most, but not quite all, of it back through disciplined use of doubling and splitting when the odds favour it, landing the residual house edge in the 0.5%–1% range on a well-behaved table.

Card counting: the shoe is not memoryless

Basic strategy assumes every card remaining in the shoe is equally likely to appear next. That is only exactly true for a single hand drawn from a freshly shuffled infinite shoe; in reality the shoe is finite and its composition changes as cards are dealt and not reshuffled every hand. Card counting exploits this: track whether the remaining shoe is relatively rich in tens and aces (good for the player — more blackjacks, better doubling and splitting outcomes, more dealer busts) or depleted of them (bad for the player), and bet and play accordingly.

The Hi-Lo system, the classic balanced count, assigns a running tag to every card seen:

2–6   →  +1   (small cards leaving the shoe favours the player)
7–9   →   0   (neutral)
10,J,Q,K,A → −1   (high cards leaving the shoe favours the dealer)

true count = running count / (decks remaining in the shoe)

Dividing by decks remaining gives the true count, which normalises the running count into a comparable measure of shoe richness regardless of how deep into the shoe you are — a running count of +6 means very different things with 6 decks left versus half a deck left. A positive true count means the remaining shoe favours the player; the standard exploitation is simply to bet more when the true count is high and less (or the table minimum) when it is low or negative, a technique called bet spreading, sometimes combined with deviating from basic strategy on specific plays (insurance becomes profitable once the true count crosses roughly +3, for instance).

How much edge counting actually buys

A skilled Hi-Lo counter with a reasonable bet spread typically converts a −0.5% house edge into roughly a +0.5% to +1.5% player edge, depending on the specific rules, penetration (how much of the shoe is dealt before reshuffling) and bet spread used — a real but modest advantage that requires betting enough money, for enough volume of hands, with disciplined bankroll management, to realise reliably rather than get wiped out by short-term variance. This is exactly what running large numbers of simulated hands is good for: expected value is a long-run average, and with only a few hundred hands the variance around it can easily swing either strategy's bankroll up or down regardless of which one is mathematically superior. Simulating a million hands is what it actually takes to watch both strategies converge cleanly to their true expected values.

Frequently asked questions

Try it live

Everything above runs in your browser — open Blackjack Strategy Simulator and change the parameters while it is running. Nothing is installed, nothing is uploaded, the whole model lives in one tab.