Sports Betting Variance — Surviving Drawdowns, Standard Deviation and Risk of Ruin
Researched and drafted with AI assistance from published sources; published by LokeNessiSport Editorial Team · Last updated: 14 August 2026
18+ | Gambling involves risk. Only bet what you can afford to lose. If gambling is causing you problems, contact BeGambleAware.org or Hjelpelinjen: 800 800 40.
Disclosure: this page contains affiliate links. If you sign up through them we may earn a commission, at no extra cost to you. It does not change what we write.
What Variance Means in Sports Betting
Variance is the gap between your short-term results and your long-term expected value. Flip a fair coin 10 times and the expected result is 5 heads and 5 tails, yet 7 heads and 3 tails is completely ordinary, and even 10 tails in a row happens about once in every 1,024 attempts.
Betting works the same way. Over a few hundred bets, luck can easily outweigh skill in either direction. Bettors who do not expect this tend to read a normal losing run as proof that their method is broken, then change it, raise stakes to win the money back, or quit just before results even out. The job of bankroll management is to make sure you can survive the swings long enough for any real edge to show.
Key takeaway: Short-term betting results are dominated by luck. Judge a method on a large sample and size stakes so that normal losing runs are survivable.
Standard Deviation: How Big Normal Swings Are
If you bet a flat stake at the same odds, your number of wins follows a binomial distribution. The standard deviation (sigma) of the number of wins over N bets is: sigma = sqrt(N * p * (1 - p)), where p is your true chance of winning each bet.
Roughly 68% of outcomes fall within one standard deviation of the expected number of wins, about 95% within two, and about 99.7% within three (the normal approximation).
Example: 1,000 bets at even money (2.00) with a true 53% chance of winning each. Expected wins: 530. Sigma = sqrt(1,000 × 0.53 × 0.47) = sqrt(249.1) = 15.78 wins. A result two standard deviations below expectation (which happens in roughly 2.3% of runs) is 530 − 31.6 = about 498 wins, a loss over 1,000 bets despite a genuine edge.
Standard Deviation and Risk of Ruin Formulas:
Advertisement · 18+ · Gambling involves risk · BeGambleAware.org
- sigma = sqrt(N × p × (1 - p))
- Z_Score = (Observed_Wins - Expected_Wins) / sigma
- Risk_of_Ruin (flat stakes at even money) ≈ ((1 - Edge) / (1 + Edge))^(Bankroll_Units)
- Edge is your expected return per unit staked; Bankroll_Units is your bankroll divided by your stake. The formula assumes even-money bets and a stake that never changes.
The Same +5% Edge at Different Odds: Expected Profit and Typical Swing Over 1,000 One-Unit Bets (arithmetic, normal approximation)
- Odds 1.50, win rate 70%: expected profit +50 units; one standard deviation ±21.7 units; chance of being behind after 1,000 bets about 1%
- Odds 1.91, win rate 55%: expected profit +50.5 units; ±30.0 units; about 5%
- Odds 2.50, win rate 42%: expected profit +50 units; ±39.0 units; about 10%
- Odds 4.00, win rate 26.25%: expected profit +50 units; ±55.7 units; about 18%
- Odds 7.00, win rate 15%: expected profit +50 units; ±79.0 units; about 26%
The same edge produces much bigger swings at longer odds, which is why longshot bettors need a larger bankroll in units, or smaller stakes, than favourite backers.
Worked example: how many bets until the edge shows? One way to ask the question is: after how many bets is the expected profit two standard deviations above zero, so that being behind becomes rare (about 2.3%)? Expected profit grows with N, while the swing grows only with √N, so the answer is N = (2 × swing per bet ÷ edge per bet)².
- At 1.91 with a 55% win rate: the swing per bet is 1.91 × √(0.55 × 0.45) ≈ 0.95 units and the edge is 0.0505, so N ≈ (2 × 0.95 ÷ 0.0505)² ≈ 1,400 bets.
- At 7.00 with a 15% win rate: the swing per bet is 7.00 × √(0.15 × 0.85) ≈ 2.50 units and the edge is 0.05, so N ≈ (2 × 2.50 ÷ 0.05)² = 10,000 bets.
Same +5% edge, seven times as many bets before the results reliably show it. Most bettors never place 1,400 bets at a consistent stake, let alone 10,000, which is why closing line value, measurable on every bet, is the more practical check.
Monte Carlo Simulation: Stress-Testing a Strategy
A Monte Carlo simulation runs thousands of simulated betting seasons with random outcomes drawn from your assumed win probabilities and odds. Instead of the single path a historical back-test gives you, it shows the range of paths your strategy could take.
From those simulated paths you can read off useful numbers: how deep a drawdown you should expect at the 95th percentile, how likely a long losing run is, and how often the bankroll would be wiped out under different stake sizes. The results are only as good as your assumed edge, so run the simulation with a smaller edge, and with no edge, as well.
A quick illustration with the risk-of-ruin formula above, for even-money bets with a 2% edge and stakes that never change: with a bankroll of 20 units (5% per bet), ((0.98) ÷ (1.02))^20 ≈ 45% chance of eventually losing the whole bankroll; with 100 units (1% per bet), ((0.98) ÷ (1.02))^100 ≈ 1.8%. Smaller stakes relative to the bankroll make ruin far less likely, even with the same edge.
- Odds Profile: Longer odds mean bigger swings, so they need more units in reserve.
- Recovery Maths: A 20% drawdown needs a 25% gain to recover; a 50% drawdown needs 100%.
- Resizing: If your bankroll falls by a set amount, for example 25%, recalculate your unit on the new total.
Advertisement · 18+ · Gambling involves risk · BeGambleAware.org
Staying Calm During Downswings
The biggest danger during a drawdown is often not the maths but the reaction to it. After a run of painful losses, such as a 96th-minute penalty or an overtime buzzer-beater, many people feel an urge to "get even".
That urge shows up in three common mistakes: (1) chasing losses by raising stakes (the Martingale idea of doubling after each loss), (2) betting on unfamiliar late-night markets, and (3) abandoning your price discipline to back short favourites just to get a win.
A simple stop-loss rule helps: for example, if you lose 4 units in one session, stop betting for 24 hours. Set the rule in advance, when you are calm.
The Circuit-Breaker Rule: Never place a bet within 30 minutes of a loss that upset you. Step away, review the bet in your log, and only resume once you feel calm.
Bankroll Protection Checklist
- Define Your Bankroll in Units: Keep your betting money separate, and divide it by 100 to set a 1-unit stake.
- Scale Stakes to the Odds: For odds above about 3.00, consider smaller stakes, for example 0.3 to 0.5 units, because swings are larger.
- Check CLV During Losing Runs: If you are consistently beating the closing price, the losses are more likely to be variance than a broken method, though that is evidence, not proof.
- Keep a Detailed Log: Record date, market, odds taken, closing price, your fair price, stake and result for every bet, and review over hundreds of bets rather than week by week.
Key takeaway: Variance is unavoidable; ruin is largely avoidable. Keep stakes small relative to your bankroll, expect bigger swings at longer odds, and judge your method on a large sample.
Staking maths does not make betting safe or profitable. Only bet money you can afford to lose; our responsible gambling page lists tools and free support. 18+.
Advertisement · 18+ · Gambling involves risk · BeGambleAware.org
Related reading
Frequently Asked Questions
What is variance in sports betting?
The gap between short-term results and long-term expected value. Over a few hundred bets luck can easily outweigh skill: with a genuine 53% edge at even money, about 2.3% of 1,000-bet runs still end with roughly 498 wins or fewer, a loss.
How do I calculate the standard deviation of my betting results?
For flat stakes at the same odds, the standard deviation of the number of wins is the square root of N x p x (1 - p). Over 1,000 bets at a 53% win rate that is about 15.8 wins; in units of profit, multiply the per-bet swing (odds x square root of p x (1 - p)) by the square root of N.
What is risk of ruin?
The chance of eventually losing the whole bankroll. For even-money bets with a 2% edge and fixed stakes, it is about 45% with 20 units in the bank (5% stakes) and about 1.8% with 100 units (1% stakes).
Why do longshot bettors have bigger swings?
Because each bet's outcome is more spread out. With the same +5% edge over 1,000 one-unit bets, the typical swing is about 30 units at odds of 1.91 but about 79 units at 7.00, and it takes roughly 10,000 bets at 7.00, against 1,400 at 1.91, before the edge reliably shows in results.