A statistical measure of how spread out results are from the average.
Standard deviation quantifies the amount of variation in gambling results. In practical terms, it tells you how much your actual results are likely to differ from the expected value.
For a series of bets, about 68% of outcomes will fall within one standard deviation of the expected value, and about 95% within two standard deviations. The higher the standard deviation, the wider the range of likely outcomes.
This concept is essential for understanding whether your results reflect skill or luck. If your actual results fall within the expected range of variance, you can't conclude much. Only when results consistently fall outside the expected range can you identify a real edge — or a real problem with your strategy.
You bet $100 flat on NFL spreads at −110 with a true 53% win rate. Expected profit per bet = +$1.43, but standard deviation per bet is roughly $99.72. Across 256 bets (one full season of ~16 bets/week), expected profit = $366, with 1-sigma range of ±$1,595.
That means a legitimately winning bettor has roughly a 32% chance of finishing the season in the red even though their edge is real. Two sigma (95% confidence) requires you to survive possible seasonal outcomes from −$2,824 to +$3,556. Sharps accept this reality and size plays accordingly — standard deviation, not win rate, dictates how much you can stake per play without risk of ruin.
<p>You bet <strong>$100 flat on NFL spreads at −110</strong> with a true 53% win rate. Expected profit per bet = <strong>+$1.43</strong>, but standard deviation per bet is roughly <strong>$99.72</strong>. Across 256 bets (one full season of ~16 bets/week), expected profit = $366, with <strong>1-sigma range of ±$1,595</strong>.</p><p>That means a legitimately winning bettor has roughly a <strong>32% chance of finishing the season in the red</strong> even though their edge is real. Two sigma (95% confidence) requires you to survive possible seasonal outcomes from <strong>−$2,824 to +$3,556</strong>. Sharps accept this reality and size plays accordingly — standard deviation, not win rate, dictates how much you can stake per play without risk of ruin.</p>
A statistical measure of how spread out results are from the average.
<p>You bet <strong>$100 flat on NFL spreads at −110</strong> with a true 53% win rate. Expected profit per bet = <strong>+$1.43</strong>, but standard deviation per bet is roughly <strong>$99.72</strong>. Across 256 bets (one full season of ~16 bets/week), expected profit = $366, with <strong>1-sigma range of ±$1,595</strong>.</p><p>That means a legitimately winning bettor has roughly a <strong>32% chance of finishing the season in the red</strong> even though their edge is real. Two sigma (95% confidence) requires you to survive possible seasonal outcomes from <strong>−$2,824 to +$3,556</strong>. Sharps accept this reality and size plays accordingly — standard deviation, not win rate, dictates how much you can stake per play without risk of ruin.</p>
Top-rated platforms reviewed by our editorial team