Bet Spreads, Bankroll and Risk of Ruin for Card Counters

By CardCount Academy · Reviewed October 5, 2026 · 11 min read

Keeping an accurate count earns nothing on its own. The money in card counting comes from one behavior: betting more when the remaining cards favor you and less when they do not. How much more is the bet spread. How much money you need to survive the swings that come with it is the bankroll question. This guide covers both, with the arithmetic shown.

Where the edge comes from

In a typical six-deck game, basic strategy leaves the house with an edge of about half a percent. With Hi-Lo, each point of true count shifts the expectation by roughly half a percent in the player's favor. Put together, as a rule of thumb:

Two things follow. First, the edge when you have it is small — one or two percent. Second, you do not have it very often. In a shoe game the true count is at or below +1 for the large majority of hands, and at +2 or higher for only a modest fraction of them, depending on penetration. A counter spends most of the time playing a slightly losing game for small bets, waiting.

What a bet spread is

The spread is the ratio between your largest and smallest bet. Betting $10 at low counts and $80 at high counts is a 1–8 spread. The reasoning is simple: if you lose a little on many small bets and win a little more on a few large bets, the total can be positive only if the large bets are large enough to outweigh the small ones.

Flat betting — the same amount every hand — does not work in a shoe game, however good the count. The many hands played at a disadvantage outweigh the few played at an advantage.

• Single and double deck: the count swings quickly and favorable situations are frequent, so narrower spreads can be enough.
• Six and eight decks: a wide spread is needed just to break even, typically 1–8 or more, or a strategy of not playing negative counts at all.

A bet ramp

A ramp is the rule that maps each true count to a bet. The teaching ramp used by our quiz and simulator is: one unit at +1 or below, two at +2, four at +3, six at +4, and eight at +5 or above. Its shape is the point — bets rise in proportion to the edge, and stay at the minimum until an edge exists.

The alternative to betting small at bad counts is not betting at all: leaving the table, or only entering a shoe once the count is positive. This is called back-counting or wonging, after Stanford Wong. It improves results more than a wider spread does, and it is also conspicuous; many casinos post "no mid-shoe entry" for exactly that reason.

Variance: why results look random

Blackjack is a high-variance game. For a flat bet of one unit, the standard deviation of a single hand is about 1.15 units, because of blackjacks, doubles and splits. When you spread your bets, the swings grow with the big bets while the edge stays small.

Take an illustrative game: a counter who averages a win of 1.5 units per 100 hands with a standard deviation of 30 units per 100 hands. Those numbers are of the right order for a 1–8 spread in a decent six-deck game, and they show the problem clearly: over 100 hands, the expected win is one twentieth of a typical swing.

After 10,000 hands — well over a hundred hours of play — the expected win is 150 units, and the standard deviation is 300 units. Being behind after 10,000 hands is not unusual; it happens to this player roughly three times in ten.

The long run has a number

A useful measure, popularized by Don Schlesinger and others, is N0 ("N-zero"): the number of hands after which your expected win equals one standard deviation. It is the variance per hand divided by the square of the win per hand.

For the example: the variance per hand is 30² ÷ 100 = 9, and the win per hand is 0.015 units. N0 = 9 ÷ 0.015² = 40,000 hands. At 80 hands an hour, that is 500 hours of play before skill is likely to show clearly over luck. Better rules and deeper penetration shrink N0 dramatically, which is why game selection matters so much.

Risk of ruin

Risk of ruin is the probability of losing an entire bankroll before it grows, if you never change your bet sizes. A standard approximation is:

Risk of ruin ≈ e^{−2 × win per hand × bankroll ÷ variance per hand}

with everything measured in betting units. For the example game:

With 400 units — $4,000 for a $10 minimum bettor spreading to $80 — this player goes broke about one time in four. Getting the risk down to a few percent takes around a thousand units. These bankrolls are much larger than most beginners expect, and they are the main reason counting is harder to profit from than to learn.

The formula assumes fixed bets forever. In practice, a player who cuts bet sizes after losing lowers the risk considerably, at the cost of earning less.

The Kelly idea

The Kelly criterion says that to grow a bankroll as fast as possible in the long run, you should bet a fraction of it roughly equal to your edge divided by the variance of the bet. In blackjack the variance of a hand is about 1.3, so with a 1% edge the full-Kelly bet is about 0.75% of the bankroll.

Full Kelly is a rough ride: large drawdowns are routine. Most serious players bet a half or a quarter of the Kelly amount, accepting slower growth for a much lower chance of a deep loss. The same idea explains the bet ramp: bets proportional to the edge.

Practical constraints

• Table limits. Your spread has to fit between the minimum and maximum of the table.
• Attention. Moving from the minimum to eight times the minimum when the count jumps is exactly what surveillance looks for. See casino countermeasures.
• Accuracy. Every figure here assumes a correct count, correct conversion and perfect basic strategy. Errors come straight out of an edge that is already thin.
• Money you can lose. A bankroll is, by definition, money whose complete loss would not change your life. If that is not true, it is not a bankroll.

What to take from this

The count tells you when; the spread decides whether it pays; the bankroll decides whether you last long enough to find out. If the numbers in this guide make counting look like a slow, uncertain grind rather than easy money, they are doing their job. Practicing it is free: try the betting quiz, and watch how a play-money bankroll moves over a few hundred hands in the simulator.

Sources and further reading

• Don Schlesinger, Blackjack Attack: Playing the Pros' Way (RGE Publishing) — origin of the "Illustrious 18".
• Peter A. Griffin, The Theory of Blackjack (Huntington Press).
• Stanford Wong, Professional Blackjack (Pi Yee Press).
• Wizard of Odds — Introduction to the High-Low Card Counting Strategy