KEY TAKEAWAYS

• A drawdown is the percentage fall from a bankroll’s peak; risk of ruin is the probability of hitting a floor such as zero.
• Each long losing streak is unlikely, but across hundreds of bets at least one is likely, even for a bettor with a genuine edge.
• Stake size sets how costly a streak is: eight straight losses cost about 15% of a bankroll at 2% per bet and about 57% at 10% per bet.
• In the hypothetical 55% example at -110, a 10-unit bankroll carries roughly a 33% modeled ruin risk versus about 0.4% at 50 units.
• The math rests on assumptions about edge, independence, and fixed stakes, so it guides stake size and never predicts results.
• Smaller stakes reduce ruin risk; raising stakes after losses makes it worse.

Risk of ruin is the probability that a bettor’s bankroll falls to zero, or to a level where continuing is no longer realistic, before any long-run edge has time to show up. A drawdown is the related, more everyday measure: the drop from a bankroll’s previous peak to its current low point. Every bettor who stays active long enough experiences drawdowns, including bettors whose estimated win probability is genuinely better than the price requires. That is why risk of ruin sits at the center of staking decisions: it connects how often losing streaks happen with how much each bet puts at stake. This article explains what drawdowns and ruin mean, shows with clearly hypothetical numbers how likely long losing streaks are, and demonstrates how stake size relative to bankroll changes both the depth of a drawdown and the chance of ruin. It also covers the assumptions behind the math, because every figure here comes from a model, not a prediction. The consistent conclusion is a cautious one: smaller stakes reduce ruin risk, and no staking pattern removes it.

What Drawdown and Risk of Ruin Mean

A drawdown is the percentage decline from a bankroll’s peak to a later low. If a $1,000 bankroll grows to $1,200 and then falls to $900, the drawdown is $300 out of $1,200, or 25%, even though the bankroll is only $100 below where it started.

The maximum drawdown is the deepest such decline over a period, and it describes how painful a stretch of results actually was. It depends on streak length, but also on how much each bet risked, which is why two bettors with identical picks can report very different drawdowns.

Risk of ruin is the probability of reaching a ruin threshold, either zero or a floor such as half the bankroll, at any point during a run of bets. Unlike a drawdown, which describes something that already happened, it is a forward-looking probability, so it needs assumptions about win probability, odds, and stake size. It is related to, but distinct from, variance in sports betting: variance describes how widely results scatter around the expected value, while ruin risk asks how likely that scatter is to push a bankroll through a floor.

Ruin does not require extraordinary bad luck. When a stake is large compared with the bankroll, an ordinary losing streak is enough. Sensible bankroll management and unit sizing exists mainly to keep ordinary streaks from becoming terminal ones.

How Losing Streaks and Stake Size Drive Ruin Risk

How Likely Is a Losing Streak?

If each bet loses with probability q and results are independent, the probability of N straight losses is qN. For a bettor who wins 55% of the time, q is 0.45, so three straight losses occur with probability 0.453 = 9.11%, five with 1.85%, eight with 0.17%, and ten with 0.034%. Each individual streak is unlikely, which is exactly why a streak feels abnormal when it happens.

The catch is opportunity. A streak can begin at any bet, and a bettor places hundreds of them. Rare events per starting point become likely across many starting points. Over 1,000 bets at those same 55% odds, the chance of at least one run of five straight losses is above 99.99%, of eight straight losses about 60%, and of ten straight losses about 17%. Over a shorter stretch of 100 bets, the figures fall to roughly 65% for five, 8% for eight, and about 1.7% for ten.

How Stake Size Sets Drawdown Depth

Streak length is outside the bettor’s control, but its cost is set by stake size. If each bet risks a fixed percentage f of the current bankroll, N straight losses leave (1 − f)N of the bankroll. After an eight-bet losing streak, that is 85.1% remaining at 2% per bet (a 14.9% drawdown), 66.3% remaining at 5% (a 33.7% drawdown), and 43.0% remaining at 10% (a 57.0% drawdown).

Recovery is lopsided too. A drawdown of d requires a gain of d ÷ (1 − d) just to return to the peak. A 14.9% decline needs about 17.5%, a 33.7% decline needs about 51%, and a 57.0% decline needs more than 130%. Deeper holes demand disproportionately larger gains, which is a large part of why ruin risk climbs faster than stake size does.

Worked Example: A Hypothetical 55% Bettor at -110

Suppose, purely as an illustration, that a bettor wins 55% of bets priced at -110 and risks one unit on each. At -110 a winning one-unit bet earns a profit of 100 ÷ 110 = 0.9091 units, and a losing bet costs 1 unit. The break-even win rate at that price is 110 ÷ (110 + 100) = 52.38%, so 55% sits above it. The expected result per bet is 0.55 × 0.9091 − 0.45 × 1 = +0.05 units. That +0.05 units of expected value per bet is an assumption for the example, not a fact about any real bettor, and these are not current sportsbook prices.

The table applies a standard continuous approximation for risk of ruin with fixed stakes: ruin probability ≈ e−2μB/σ², where μ is the expected profit per bet (0.05), σ² is the variance of one bet’s result (about 0.9020), and B is the bankroll measured in units. Each bet risks one unit, so the bankroll size also fixes the stake as a share of the bankroll.

Bankroll (units) One-unit bet as % of bankroll Approximate risk of ruin
10 10% 33.0%
20 5% 10.9%
50 2% 0.39%
100 1% 0.0015%

Read across the rows, a bigger cushion cuts ruin risk far faster than proportionally: moving from 10 to 20 units drops the modeled figure from about 33% to about 11%, and 50 units pushes it below 0.4%. The same positive edge is present in every row. Only the stake relative to the bankroll changed.

How to Read the Risk of Ruin Formula and Its Limits

In the formula, the exponent grows with the edge and the bankroll. A larger μ or B makes the negative exponent bigger and ruin less likely, while a larger σ², as with long-shot prices, makes ruin more likely. In plain terms, ruin risk falls exponentially as the cushion grows, which is what the table shows.

Every input is an assumption, and the edge is estimated, not known. If the true win probability at -110 were 52.38% instead of 55%, the expected result would be zero, and over an unlimited horizon ruin becomes a matter of time rather than chance. Because μ sits directly in the exponent, small errors in the estimated edge move the result a lot, which is one reason a small sample of results cannot confirm an edge.

The formula also assumes independent bets, fixed stakes, and a known price. Real results can cluster, for instance when parlay legs are correlated, stakes change, and lines move. Treat the output as an order-of-magnitude guide to how stake size shifts risk, not as a probability to quote to the decimal.

Common Mistakes and Misconceptions

“A positive edge means no ruin risk” is the most common error. Positive expected value lowers ruin risk but never removes it, and oversized stakes can make ruin likely even with a real edge. The 10-unit row above shows roughly a one-in-three modeled chance.

Believing a streak is “due” to end is the gambler’s fallacy in another form. After eight losses, the next bet still carries the same 55% win probability under the model’s independence assumption. Streaks end because of the odds on each bet, not because results owe the bettor a correction.

Raising stakes after a drawdown is the most dangerous mistake. A deeper hole needs a larger percentage gain to recover, and larger stakes shrink the number of losses a bankroll can absorb, so ruin risk rises at exactly the wrong moment. Chasing losses is a pattern to avoid, never a recovery plan.

Using Ruin Risk to Think About Stake Size

In practice, risk of ruin turns a vague question about how much is too much per bet into a concrete one: how many losing bets in a row can the bankroll absorb? A bettor can estimate the longest streak that is plausible over the number of bets they expect to place, then check how deep the drawdown would be at their current stake. If an eight-bet streak would take more than a third of the bankroll, as in the 5% case above, the stake is a bigger commitment than it first appears.

The comparison of flat and percentage staking is useful here. Percentage staking shrinks stakes as the bankroll falls, so in this simple model it never reaches exactly zero, while flat staking can. Neither removes drawdowns. A bankroll should only ever be money that is not needed for essential expenses, and a bettor who finds drawdowns stressful is showing that stakes are too large for their situation, whatever the math says.

Risk of ruin is the downside half of a larger staking picture. The Kelly Criterion takes an estimated edge and price and returns a stake fraction that maximizes theoretical long-run growth. Full Kelly sizing can produce deep drawdowns and is very sensitive to an overestimated edge, which is why fractions of it are often discussed instead. Reading it alongside the ruin math here shows why the cautious direction is almost always smaller.

The practical takeaway is simple. Decide the stake from the bankroll’s capacity to absorb a long losing streak, not from confidence in the next pick, and never raise it to win back losses. Results will still vary, and no stake size makes a bet certain to win.

Frequently Asked Questions

What does risk of ruin mean in sports betting?

Risk of ruin is the probability that a bankroll falls to zero, or to a chosen floor, at some point during a series of bets. It is a modeled estimate based on assumed win probability, odds, and stake size, so it describes risk under stated assumptions rather than predicting any bettor’s results.

What is a drawdown in sports betting?

A drawdown is the percentage decline from a bankroll’s highest point to a later low. If a bankroll climbs to $1,200 and falls to $900, the drawdown is 25%. The maximum drawdown is the deepest decline over a period, and it grows with both losing streak length and stake size.

Can a bettor with an edge still have a long losing streak?

Yes. A bettor who wins 55% of bets still faces a 0.17% chance of eight straight losses on any given starting bet, but across 1,000 bets the chance of at least one such run is about 60%. An edge changes long-run expectation, not the certainty of streaks.

What is the probability of ruin?

It depends on the edge, the price, and stake size relative to the bankroll. In a hypothetical 55% example at -110 with one-unit bets, the modeled figure is about 33% for a 10-unit bankroll and under 0.4% for 50 units. With no edge, ruin becomes a matter of time.

Is risking 5% of a bankroll on each bet too much?

There is no universal answer, but the numbers are sobering. Eight straight losses at 5% per bet leave about 66% of the bankroll, a 33.7% drawdown, and recovery needs roughly a 51% gain. A smaller stake reduces both the depth of drawdowns and the chance of ruin.

How does stake size change risk of ruin?

Larger stakes relative to the bankroll mean fewer losses can be absorbed, so ruin risk rises quickly. In the hypothetical example, a 10% stake gives about a 33% modeled ruin risk, while a 2% stake gives about 0.4%. Smaller stakes reduce ruin risk but never eliminate it.