The Question Sits on a Misread Story
Most versions of the Don Johnson blackjack story compress a months-long negotiation into a single image: one man, a stack of chips, and a counting method that supposedly cracked the game. That framing leads straight to the wrong question. The interesting part of the story is not which count he used. It is that the terms of the game itself were changed before the first card was dealt, and those changed terms are what moved the arithmetic in his favour.
So the honest answer to the query is: no published count, on its own, explains what happened. A counting system operates on the game as it is dealt. What Johnson secured was a game dealt on different terms. Those are two different levers, and confusing them is how a story about negotiation becomes a story about a magic system.
What a Card Count Measures
Card counting is a way of tracking the composition of the remaining shoe. When low cards leave the deck, the remaining shoe is richer in tens and aces. That shift changes the probability of blackjacks, of stiff hands, and of dealer busts. A counter raises the size of a wager when the remaining composition favours the player, and lowers it when it does not.
The count does not change the rules of the game. It does not change the payout on a blackjack, the dealer's standing rules, or the number of decks in the shoe. It changes only the player's information, and therefore the size of the bet placed at a given moment. The house edge on any single hand is still built into the rules. What the counter is doing is betting more when the conditional edge has shifted, and less when it has not.
That shift is real but small. A counter playing a standard multi-deck shoe with typical rules is not turning a losing game into a winning one by a wide margin. The edge moves by fractions of a percentage point per true count, and it only exists on some fraction of the hands dealt. The rest of the time the counter is sitting at the same negative expectation as everyone else.
The Terms Johnson Negotiated Changed the Arithmetic
The reported details of the Johnson story matter here because they are structural, not personal. The accounts describe a set of negotiated concessions: a discount on losses, a higher table limit than the posted one, and rules that reduced the house's built-in advantage. Each of those concessions is a change to the game's parameters, not a change to how the cards were read.
A loss discount works like this. Suppose a player and the venue agree that after a session, a percentage of the player's losses will be returned. That return is applied to the total lost, not to each hand. It does not change the probability of any hand. It changes the settlement after the fact. If a player loses a large amount, the discount returns part of it. If the player wins, the discount does not apply. The effect is to tilt the session-level outcome in the player's favour without touching the hand-level edge.
A higher table limit changes the size of the swings available. It does not change the edge, but it changes how much of the player's bankroll can be put at risk on a single favourable situation. When combined with a count, a higher limit lets a player bet more when the remaining shoe is rich in tens and aces. Without the changed terms, the same count produces much smaller absolute results.
The rule concessions are the most direct change to the edge itself. Blackjack's house edge is a function of specific rules: how many decks are in the shoe, whether the dealer stands or hits on soft totals, whether doubling after splitting is allowed, whether surrender is offered. Each rule change moves the edge by a small amount. A player who negotiates several favourable rule changes is playing a different game than the one on the posted layout. The count is then applied to a game that already has less house edge built in.
Expected Value Under the Two Levers
Expected value is the long-run average result of a bet, given the probabilities and payouts. For a blackjack hand, the expected value is negative for the player under standard rules. Card counting shifts the expected value of some hands from negative to positive, but only when the remaining shoe is favourable, and only by a small amount. Negotiated rule changes shift the expected value of every hand, regardless of the count.
Here is the arithmetic as an example, with hypothetical inputs. Suppose a game has a house edge of one unit per hundred wagered under its posted rules. A counter might move that to roughly break-even on average across all hands, by betting more when the count is favourable and less when it is not. Now suppose the same game is dealt with a rule concession that removes a quarter of the house edge. The same counter is now playing a game where the baseline edge is smaller, and the favourable-count hands are worth more. The two effects multiply, but they are not the same effect.
The distinction matters because a reader who asks which count Johnson used is usually asking for the count itself, as if the count were the engine. The engine in the story was the renegotiated game. The count, if any was used, was the transmission. A transmission does not move a car without an engine.
Why Short Sessions Look Nothing Like the Long Run
Expected value is a long-run average. A single session, even a long one, is a small sample. In a game with variance as high as blackjack, the actual result of a session can be far from the expected value in either direction. A player with a small positive edge can lose for a long time. A player with a negative edge can win for a long time. The gap between the two is variance, and variance dominates short samples.
This is why a single winning session proves nothing about a system, and a single losing session disproves nothing about a count. The only way to observe an edge is over enough hands that the variance shrinks relative to the expected value. How many hands that takes depends on the size of the edge and the size of the variance. A small edge takes a very large number of hands to distinguish from zero.
In the Johnson story, the reported sessions were high-stakes and relatively short in hand count. The negotiated terms, particularly the loss discount, were designed to work at the session level, where variance is large. A loss discount does not require a long run to matter. It applies to whatever was lost, whenever the session ends. That is a different kind of edge than a counting edge, and it is visible in a single session in a way a counting edge is not.
Decisions Per Hour and Total Exposure
The number of hands played per hour changes the total amount wagered, and therefore the total expected loss or gain. A player at a full table might see far fewer hands per hour than a player heads-up. A player who bets more per hand, or who plays faster, is putting more money through the game in the same time.
Total expected value over a session is the expected value per hand multiplied by the number of hands, multiplied by the average bet. If the per-hand edge is negative, more hands and larger bets mean a larger expected loss. If the per-hand edge is positive, more hands and larger bets mean a larger expected gain. The count changes the average bet over time, but it does not change the number of hands dealt. The negotiated table terms changed the per-hand edge itself.
This is why the high table limit mattered in the story. A player with a small edge per hand needs a large bet to make the edge meaningful in absolute terms. A player with a small edge per hand and a small bet is earning very little per hour, even when the count is favourable. The limit is not the edge. It is the multiplier on the edge.
Progression Systems Redistribute, They Do Not Remove
A progression system changes the size of the next bet based on the previous outcome. The most common version doubles the bet after a loss, so that a win recovers the accumulated losses plus one unit. The arithmetic looks appealing on paper, but it does not change the per-hand edge. It changes only the distribution of outcomes.
Here is the mechanism as an example. Suppose a player starts with a one-unit bet and doubles after each loss. The sequence of bets is one, two, four, eight, and so on. A win at any point recovers all prior losses plus one unit. The problem is that the player is not guaranteed a win before the bet size reaches the table limit or the bankroll runs out. When the losing streak ends beyond the affordable bet size, the loss is much larger than any single unit. The progression has concentrated the losses into rare, large events, and the wins into frequent, small events. The expected value is unchanged. The variance is higher.
Card counting is not a progression system. A counter changes the bet based on the composition of the remaining shoe, not on the previous hand's outcome. But the two are often confused because both involve changing bet size. The distinction is that the count carries information about the next hand's conditional probability, while a progression carries no information at all. The count can shift the conditional edge. The progression cannot.
The Johnson story is sometimes retold as if a progression or a bold betting style broke the casino. That retelling misses the structural point. The terms of the game were changed, and the betting was sized to those terms. No betting pattern, however bold, removes a negative expectation from an unchanged game.
What the Reader Can Take From the Story
The useful lesson is not which count to learn. It is that blackjack's edge is a function of the rules, and the rules are sometimes negotiable for a player with enough at stake. The posted layout is not always the only game available. A player who understands which rule changes move the edge, and by how much, is in a position to ask for them. A player who only learns a count is playing the game as posted, and the posted game is negative.
The second lesson is that the count and the terms are separate levers. A count works on the information available during play. The terms work on the game itself. Both can be present, both can be absent, and neither substitutes for the other. A player with favourable terms and no count is playing a better game than a player with a count and standard terms, because the terms apply to every hand and the count applies only to some.
The third lesson is that variance hides everything in the short run. A single session, even a famous one, is not evidence of an edge. The only way to know whether a method works is to compute the expected value under the actual rules, and the only way to observe it is over enough hands that the variance shrinks. The Johnson story is a story about a negotiated edge, not a counting edge, and the difference is the whole point.
FAQ
Did Don Johnson use a specific published card counting system?
The public accounts of the story do not identify a specific published counting system as the source of his results. The reported concessions, including a loss discount and rule changes, were negotiated before play began. Any counting that occurred would have operated on top of those changed terms, not instead of them.
Can card counting overcome the house edge in blackjack?
Card counting can shift the conditional edge on some hands from negative to positive, but the shift is small and only present when the remaining shoe is favourable. Under standard rules, the overall edge across all hands is still close to break-even or slightly negative for most counters. The published basic-strategy charts and rule-specific edge figures are the starting point for any such calculation.
Why do casinos negotiate different rules with high-stakes players?
A venue may negotiate rules, limits, or loss discounts with a player whose expected total wager is large enough to justify the concession. The concession changes the game's expected value for that player, but it does not change the underlying probabilities of the cards. The venue is trading a smaller edge per unit wagered for a larger total amount wagered.
Play responsibly. If you or someone you know has a gambling problem, call the helpline printed in the footer below.
Play responsibly
Gambling should be treated as paid entertainment, never as a way to earn income or recover losses. 1-800-MY-RESET (1-800-697-3738)
18+ or 21+ depending on where you are; follow the minimum age that applies to you.
This article is general information about how these mechanics work. It is not legal advice and not a recommendation to gamble or to use any particular operator. Availability and legality differ by jurisdiction — check the rules that apply where you are.