Risk of Ruin in Blackjack: How to Calculate It and Size Your Bankroll

10 min read · October 22, 2026

Risk of ruin (RoR) is the probability that you lose your whole bankroll before you reach a goal or before your edge has time to pay off. In blackjack it depends on three things: your edge, how much a single hand swings your results, and how many betting units you hold. A basic strategy player has no edge, so ruin is only a matter of time. A card counter has a small edge, and a bankroll of roughly 200 to 400 units usually keeps ruin low.

This guide shows you the formula, works through real numbers, and explains what to change when the result scares you.

TL;DR

Risk of ruin is the chance of losing your entire bankroll. It falls as your edge or your bankroll grows, and rises as your bets get bigger relative to your bankroll.
A simple estimate is RoR = e^(-2 × edge × units / variance). With a 1% edge and about 1.32 variance, 200 units gives roughly a 5% risk and 300 units roughly 1%.
Card counters are commonly advised to hold 200 to 400 betting units. Fewer units means a much higher chance of going broke.
Without an edge, there is no safe bankroll. Basic strategy at a typical table only slows the loss.
Cutting your unit size is the fastest, cheapest way to lower risk of ruin.

What risk of ruin actually measures
Why blackjack is a high-variance game
How to calculate risk of ruin
How many units do you need?
Session risk is not the same as total risk
What if you do not count cards?
Betting systems do not lower your risk
Kelly betting and why most counters go fractional
Mistakes that raise your risk of ruin
Our perspective
Practice before you put money at risk
A short checklist before you play
FAQ
Sources

What risk of ruin actually measures

People often read RoR as “the chance I lose tonight.” It is a different idea. Risk of ruin is the chance that a bankroll of a given size hits zero if you keep playing the same way, usually with no time limit. Some sources add a goal. A common version asks the chance of losing your bankroll before you double it, or before you hit a profit target.

The distinction matters because the answer changes with the question you ask:

Ruin with no limit: you play forever. A counter with a real edge can still go broke, but the odds shrink quickly as the bankroll grows.
Ruin over a trip or session: you play a fixed number of hours. The risk is smaller, because there is less time for a bad run.
Ruin before doubling: you stop at a win goal. This is the version many magazine articles use.

If you have been reading about blackjack variance, RoR is where variance becomes a number you can plan around.

Why blackjack is a high-variance game

Even a winning player spends a lot of time losing. The standard deviation of one hand in standard blackjack is about 1.15 bets, which means a single hand regularly moves your result by more than your average expected gain over hundreds of hands. Your edge is tiny next to that noise.

The math behind this is old. The classic “gambler’s ruin” problem shows that even a slightly negative expectation does serious harm over time, an effect many people assume is harmless. Edward Thorp’s central point on this was that finding an edge is only half the job. Sizing bets so you survive long enough to collect on it is the other half.

How to calculate risk of ruin

The usual quick formula is:

RoR = e^(-2 × Edge × Units / Variance)

This formula appears on Practical Web Tools with the same inputs. Here is what each piece means:

Edge is your expected profit per unit bet, as a decimal. A 1% edge is 0.01.
Units is your bankroll divided by your average bet. A $5,000 bankroll with a $25 average bet is 200 units.
Variance is the square of the standard deviation. With a standard deviation of 1.15, variance is about 1.32.
e is the constant 2.718, which your phone calculator has as the exp button.

This is an approximation. It treats every hand as one bet with the same edge, while real counters vary their bets with the count, double, split and take blackjacks at 3:2. For a careful answer, serious players use simulation software. For planning, the formula is close enough to show you the shape of the problem.

A worked example

Say your average edge is 1%, and variance is 1.32. The exponent is -2 × 0.01 × units / 1.32, or about -0.01515 per unit. The table below uses that figure. These are our own calculations from the formula above, not figures from a casino study.

Bankroll in units	Exponent	Risk of ruin (approx.)
50	-0.76	47%
100	-1.52	22%
150	-2.27	10%
200	-3.03	5%
300	-4.55	1%
400	-6.06	0.2%

Notice the curve. Going from 100 to 200 units cuts the risk from about 22% to about 5%. Going from 300 to 400 units buys much less. Past a point, extra bankroll protects you from a danger that is already small.

The same bankroll with a thinner edge

Now assume your average edge is only 0.5%, which is realistic for a counter who plays a mediocre game or spreads bets narrowly. At 200 units the exponent is -2 × 0.005 × 200 / 1.32, about -1.52, and the risk is about 22%. The same bankroll that was a comfortable 5% at a 1% edge is now a roughly one-in-five chance of ruin. Halve the edge and you need roughly double the units to hold the risk steady.

Risk of ruin rises as blackjack edge falls

This is why table rules and penetration matter so much. They change your edge, and your edge is the biggest lever in the formula.

How many units do you need?

Practitioners cluster around the same range. Professional recommendations for card counters generally run from 200 to 400 betting units, and beginners tend to underestimate how much that is. At a $25 unit, 200 units is $5,000 and 400 units is $10,000.

Other sources give very different numbers because they assume different bet spreads. One counting guide puts a $25-unit counter at about $15,000 for a 10% risk of ruin. That is 600 units, well beyond the range above. The lesson is not that one source is wrong. Your own spread, penetration and rules decide your edge and variance, so you have to run your own numbers.

Here is a rough way to turn a target into a unit size:

Decide how much money you can lose without it affecting rent, bills or savings. That is your bankroll.
Pick a risk of ruin you can live with. Many players aim for 1% to 5%. Higher than 10% is a gamble by any definition.
Find the units that match, using the formula or the table above.
Divide the bankroll by the units. That is your unit size.

A $6,000 bankroll at 300 units gives a $20 unit. If the only table nearby has a $25 minimum, you are underbankrolled for that table and should either find a cheaper game or accept more risk. Our guide to table minimums and bet size covers how that mismatch plays out.

Session risk is not the same as total risk

The no-limit figure is the scary one. A single visit is gentler. One estimate for a 0.5% house-edge game at $10 a hand says a $200 bankroll has about a 1% chance of ruin in a one-hour session, and a $400 bankroll about 3% over four hours. Those figures are for a player facing the house edge, not a counter, and they show something useful: the longer you sit, the more bankroll you need just to survive the session.

Short bankrolls also look worse than they feel. One magazine example shows a player who goes from a $5 to a $10 bet with a $200 bankroll. That halves the bank to 20 units and lifts the risk of ruin to 40%, meaning four out of ten four-hour sessions end in a busted $200. Same money, same game, double the bet.

What if you do not count cards?

A basic strategy player at a typical table gives up a small percentage to the house on every hand. There is no positive edge for the formula to work with, so every bankroll eventually goes to zero. The question becomes how fast, and how much entertainment you buy per dollar.

One simple illustration comes from the same magazine piece. It assumes a 0.5% house edge and says a basic strategy player has about a 46% chance of doubling a $200 bankroll at $10 units before losing it. That is a coin flip with a slight tilt against you, over a session with no time limit.

If you play for fun, treat your bankroll as a cost, not an investment. Set a loss limit before you sit down, and walk when you hit it. If you want to cut the house edge, start with basic strategy and a table with better rules. Fixing mistakes is the cheapest edge you can buy.

Betting systems do not lower your risk

The Martingale is the classic trap. Doubling after every loss often produces a winning session, which feels like proof. Experts call it a mathematical illusion: it makes a catastrophic wipeout in one long losing streak far more likely. Table limits and a finite bankroll make the problem worse. We explain the arithmetic in why doubling bets does not work.

The formula tells you why. Risk of ruin depends on edge, variance and units. Changing the order of your bets changes none of those.

Kelly betting and why most counters go fractional

The Kelly Criterion tells you what fraction of your bankroll to bet to maximize long-run growth. In theory, it is the gold standard. In practice, full Kelly is too aggressive for most people, because it produces big drawdowns that can break a player emotionally before the math pays off. Most professional counters use a fraction, commonly half Kelly, which gives up a modest share of growth for a much smaller chance of deep drawdowns.

The same source describes two naive answers that both fail. Betting everything each round makes ruin almost certain, even with a small edge. Betting too little means you barely grow. Kelly sits between them, and half Kelly sits a bit more safely on the cautious side.

For a counter, this shows up as your bet spread. You bet more when the true count is high and less, down to the minimum, when it is low. Your spread, not your intentions, sets your real risk. See our guide on card counting deviations for how high counts also change your plays.

Mistakes that raise your risk of ruin

Overbetting. Betting a larger unit than your bankroll supports is the most common way counters go broke, even when their long-term edge is real.
No stop-loss. Practitioners note that counters bet aggressively at a high true count but must still stick to a session loss limit to survive normal negative swings.
Counting errors. A wrong running count makes you bet big when you have no edge. Your formula inputs assume you play accurately.
Playing a weak game. Poor penetration, a bad payout or a high minimum lowers your edge and raises your required units.
Chasing losses. Raising your unit after a bad run to “get back to even” shrinks your units at the worst moment.

Our perspective

Most readers who search for risk of ruin have already learned something about counting, and the formula is where it becomes real. In building TwentyOne, we have seen the same pattern: people practice the count and the strategy, then size their bankroll by feel. That is backwards. The math you practice only pays if you survive the swings.

Our view is simple. First, get your play accurate, because every error comes out of your edge, and a thinner edge means a higher risk of ruin. Second, size your unit from your bankroll, not from the table minimum you want to play. Third, run the numbers for the specific rules you will face. A game with a different penetration, number of decks or dealer rule is a different bankroll.

That last point is why the app lets you customize the rules to match a real casino, and why we put instant feedback on every hand. Accuracy under real conditions is what protects your edge. The formula is a planning tool. It will not make a bad plan safe, and it will not promise you a result. No bankroll removes risk, and no one should play with money they cannot afford to lose.

— Ryan

Practice before you put money at risk

A bankroll is easiest to protect when you have already made your mistakes for free. TwentyOne is an iPhone trainer with lessons on basic strategy and card counting, drills for the count, quizzes, and instant feedback on every hand explaining why a play is right or wrong. You can set the rules to match the casino you plan to visit, and track your performance as you improve. Start with the free lessons in the TwentyOne app, and use the interactive strategy chart to check plays for your table’s rules. If you are building a counting routine, our drill plan for practicing at home shows how to structure it. Premium features and a lifetime option are available through the App Store.

A short checklist before you play

Write down your bankroll, the money you can afford to lose entirely.
Estimate your edge honestly for the rules and penetration you will actually get.
Pick a target risk of ruin, then find the units it needs.
Divide to find your unit, and check it against the table minimum.
Set a session loss limit and stick to it.
Review the numbers after each trip. If your results run well below expectation, check your accuracy first, then your assumptions.

For the wider picture on managing money at the table, see our full bankroll management guide. If you want to know what happens once you start winning, read about counting cards in a casino.

FAQ

How do you calculate risk of ruin in blackjack?

Use RoR = e^(-2 × Edge × Units / Variance). Enter your edge as a decimal, your bankroll divided by your average bet as units, and about 1.32 for variance. It is an approximation, so for exact figures use a simulation with your own rules and bet spread.

What is a good risk of ruin for blackjack?

Many card counters aim for 1% to 5%, which at a 1% edge means roughly 200 to 300 units by our calculation. A risk near 10% or higher means you are likely to go broke in a meaningful share of long runs. The right figure depends on how much you can lose without harm.

Can basic strategy players use risk of ruin?

The formula assumes a positive edge, so it does not help at a normal table where the house has the advantage. There, you will lose over time, and the useful question is how much you are willing to spend. Set a loss limit and treat it as entertainment cost.

Does a bigger bankroll remove the risk?

No. A bigger bankroll lowers the risk, and the curve flattens as units grow, but it never reaches zero. Bad variance still happens, and counting errors or a weak game can erase the benefit.