Risk of ruin starts with a bankroll that can run out
Most blackjack players think about a session as a series of hands. A risk-of-ruin calculation thinks about it as a single question: given this starting bankroll, this bet size, and this game's built-in edge, what are the chances the money reaches zero before it reaches any other target? The answer is always a probability between 0 and 1, and it is never zero for a player facing a negative-expectation game.
The calculation treats your bankroll as a path. Each hand moves the path up or down by the amount you wagered, adjusted for wins and losses. Ruin is the event where the path touches zero. The formula that describes this path is not a casino secret; it is a standard result from probability theory, and it needs only a few inputs: your starting bankroll, your average bet, the house edge on the game you are playing, and the variance of that game.
What the calculation needs from you
Before any arithmetic, you need three numbers that describe your specific situation, not blackjack in general. The first is your bankroll, expressed in the same units as your bet. The second is your average bet per hand. The third is the edge on the specific rule set at your table. Blackjack edges differ materially by rule set: the number of decks, whether the dealer stands or hits soft 17, whether doubling after splitting is allowed, and the payout on a natural all move the edge. A published basic-strategy chart only applies to the rule set it was built for, so the edge you use must come from the same rule set you are playing.
The fourth input is variance, which describes how much a single hand's result swings around its average. Blackjack is not a coin flip. A natural pays more than a standard win, splits and doubles create multiple units at risk on one hand, and pushes return the stake without profit. All of that raises variance relative to a game where every wager wins or loses one unit. A risk-of-ruin calculation that ignores variance will understate ruin probability for a real blackjack player.
| Input | What it represents | Where the player gets it |
|---|---|---|
| Bankroll | Total funds set aside for blackjack, in betting units | The player's own budget decision |
| Average bet | Typical wager per hand, including doubles and splits | The player's own bet sizing |
| House edge | Theoretical percentage of each wager the game retains | Published figures for the exact rule set in play |
| Variance | How widely individual hand results swing | Derived from the game's payout structure and rules |
If any of these inputs is a guess, the output is a guess. The method is reliable; the result is only as good as the numbers fed into it.
The arithmetic, step by step
Risk of ruin is usually expressed through a formula built on the logarithm of wealth, because a bankroll grows and shrinks multiplicatively rather than additively. For a simplified game with a fixed bet and a known edge, the ruin probability for a player who plays indefinitely is approximately the exponential of a negative quantity: the bankroll divided by the bet, multiplied by twice the edge, divided by the variance. That is a mouthful, so here is the structure in labelled pieces.
Suppose a player brings 200 units to a table and bets one unit per hand. Suppose the rule set in play gives the house an edge of 0.5%, expressed as 0.005. Suppose the variance per hand is 1.3, a reasonable order of magnitude for blackjack with doubles and splits. The exponent in the ruin formula is negative two times the edge times the bankroll, divided by the variance, all divided by the bet. Plugging in the example numbers: two times 0.005 times 200 is 2.0. Dividing by 1.3 gives about 1.54. The negative of that is about -1.54. Raising the mathematical constant e to that power gives roughly 0.21. In this example, the long-run risk of ruin is about 21%.
That number is an example built from hypothetical inputs, and it is marked as such. The point is not the specific percentage. The point is the shape of the relationship. Double the bankroll while keeping the bet fixed, and the exponent doubles in magnitude, which drives the ruin probability down sharply. Double the bet while keeping the bankroll fixed, and the exponent halves, pushing ruin probability up. Raise the house edge, and ruin probability rises. Raise variance, and ruin probability rises even with the same edge, because wider swings make it easier for a downswing to reach zero before an upswing restores the bankroll.
Why a short session looks nothing like the long-run number
A risk-of-ruin figure is a long-run probability for a player who keeps playing. It is not a prediction for tonight's session. Over a small number of hands, variance dominates. A player with a low long-run ruin probability can easily win for an evening, and a player with an even lower ruin probability can lose an entire bankroll in a single bad run. The calculation does not tell you what will happen next hand; it tells you what fraction of imagined parallel sessions end in ruin if the same strategy runs forever.
Decisions per hour matter here for a reason most players overlook. Each hand is an independent exposure to the house edge. The more hands you play, the more total wager you put through the game, and the more the theoretical edge accumulates in absolute terms. A player betting one unit per hand at a 0.A five percent edge loses an expected small fraction of coin-in.005 units per hand. At 60 hands per hour, that is 0.3 units of expected loss per hour. At 120 hands per hour, it is 0.6. The per-hand edge did not change; the total exposure did. Risk of ruin responds to total exposure, so a faster game or a longer session raises the probability of touching zero within a fixed bankroll.
Progression systems do not remove the ruin risk
A common instinct when facing a losing streak is to raise the bet to win back losses faster. This is a progression system, and it changes the risk-of-ruin calculation in a predictable way: it raises variance without changing the house edge. The edge is built into each wager independently. Doubling a bet after a loss doubles the expected loss on that next hand, because the edge applies to the larger amount. What the progression does is redistribute outcomes: it creates many small wins and occasional large losses, instead of a smoother path. The long-run ruin probability does not improve, and in many cases it worsens, because the larger bets accelerate the path toward zero during a downswing.
What genuinely changes the inputs is bet selection and rule set. A player who chooses a table where the dealer stands on soft 17, where doubling after splitting is allowed, and where a natural pays the better rate faces a lower edge than a player at a table with worse rules, assuming identical basic strategy. Lower edge lowers ruin probability for the same bankroll and bet. Slower play lowers total exposure per hour. A smaller bet relative to bankroll lowers ruin probability. Those are the levers the calculation exposes. The progression is not one of them.
How to check a published risk-of-ruin figure
Many calculators and charts circulate for blackjack risk of ruin, and they are not all built on the same assumptions. Some assume a fixed bet; some assume a bet spread. Some assume perfect basic strategy; some assume a card-counting system with a specified edge. Some use variance figures for a single-deck game; some use multi-deck figures. Before trusting any published number, check what edge and variance the source used, whether the bet is fixed or variable, and whether the result is a long-run probability or a session-specific simulation. A figure without those inputs stated is not verifiable.
For a player who wants the number for their own situation, the reliable route is to compute it from the formula with their own bankroll, bet, and the edge for the exact rule set they play. Publicly published reference sources list edges for common rule combinations, and those figures change as rule sets change. The calculation itself is straightforward once the inputs are pinned down. The hard part is being honest about the inputs, especially the edge, because the edge is the one input the player cannot observe at the table. It must come from the rules, not from how the last hour felt.
FAQ
Does risk of ruin apply to a single blackjack session?
It can be adapted to a session, but the standard formula describes long-run play with a fixed strategy. A session-specific ruin probability depends on the number of hands you intend to play, your bet, and the game's variance, and it is usually estimated by simulation rather than the closed-form long-run formula. The long-run number still matters because it tells you what the strategy does if you keep returning to the table.
Can card counting change the risk of ruin calculation?
Card counting changes the edge input. If a counting system gives the player a positive edge at certain counts, the same ruin formula applies with that edge substituted in, and the ruin probability drops for a given bankroll and bet. The variance input also changes because bet spreads and larger wagers at high counts widen the swings. The calculation is the same; the inputs are different.
Why do two risk-of-ruin calculators give different answers for the same bankroll?
They are almost certainly using different assumptions. One may assume a fixed bet while another assumes a spread, one may use a different variance figure, and one may be reporting a session probability while another reports a long-run probability. Compare the stated edge, variance, and bet model before comparing the output. A calculator that does not state its inputs is a black box, and its output cannot be checked.
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.