Definition
Variance describes the natural statistical swings around expected value in betting results. Even a 5% edge does not prevent long losing streaks — a bettor with 55% win rate at −110 will lose 5 or more consecutive bets roughly 1.5% of the time, which is multiple times per season. Variance is unavoidable and not a signal of a broken system. Higher-odds bets (underdogs, parlays) have higher variance. Larger sample sizes reduce variance's distortion of true ROI.
Worked Example
True edge: 55% win rate at −110. Standard deviation per bet ≈ $104. Over 100 bets at $110 each: expected profit = (0.55 × $100 − 0.45 × $110) × 100 = $550 − $4,950... Corrected: (55 × $100) − (45 × $110) = $5,500 − $4,950 = $550. 1 standard deviation band: ±$104 × √100 = ±$1,040. Realistic range: −$490 to +$1,590 while having a real edge.
Why It Matters
Understanding variance prevents abandoning a winning process during a losing streak. A 10-game losing run at 55% true win rate happens — it means nothing about your system's validity.