How to Calculate Expected Value in Sports Betting (With Real Examples)
Learn how to calculate expected value in sports betting with the EV formula, real examples, and tips to find +EV picks before you place a wager.
Expected value is the single most important concept separating disciplined sports bettors from casual gamblers—and once you understand the math, you'll never look at a betting line the same way again. In this guide you'll learn exactly how to calculate EV, see it applied to real wagers, and understand how to use it as a filter before every pick you consider.
What Is Expected Value in Sports Betting?
Expected value (EV) is a mathematical concept borrowed from probability theory. In sports betting, it measures how much you can expect to gain or lose on average for every dollar wagered, assuming a given set of probabilities holds over a large sample. A positive expected value (+EV) means the bet is mathematically favorable; a negative expected value (−EV) means the house edge works against you.
The key insight is this: sportsbooks set lines to reflect their own margin (the "vig" or "juice"), which systematically pushes most bets into −EV territory. Finding spots where the true probability of an outcome is higher than the implied probability baked into the odds is where long-term edge lives. You are not trying to win every bet—you are trying to make decisions that are correct on average.
The Expected Value Formula Explained
The standard EV formula for a binary sports bet (win or lose) is:
EV = (Probability of Winning × Amount Won per bet) − (Probability of Losing × Amount Lost per bet)
To use it, you need two inputs: (1) the payout implied by the odds, and (2) your own estimated true probability of the outcome. The gap between your probability estimate and the sportsbook's implied probability is your potential edge.
- Convert American odds to implied probability: For +150 odds → 100 ÷ (150 + 100) = 40.0% implied probability.
- For negative odds (−120) → 120 ÷ (120 + 100) = 54.5% implied probability.
- Subtract the vig to get the "break-even" probability the book needs you to lose at.
- If your estimated true probability exceeds that break-even threshold, the bet is +EV.
A Step-by-Step EV Calculation Example
Suppose a sportsbook prices an NFL underdog at +200. The implied probability is 100 ÷ (200 + 100) = 33.3%. Now suppose your own analysis—using injury reports, line movement, situational angles, and historical splits—leads you to estimate the underdog's true win probability at 42%.
Plug into the formula (assuming a $100 wager, where a win pays $200):
EV = (0.42 × $200) − (0.58 × $100) = $84 − $58 = +$26
That +$26 figure means that, on average, every time this exact scenario repeats at these odds and this true probability, you expect to come out $26 ahead per $100 wagered. That is a strongly +EV bet. Conversely, if your true probability estimate were only 30%—below the book's implied 33.3%—the EV would be negative and the bet should be passed.
- Always calculate EV before acting, not after—post-result rationalization is one of the most common analytical errors.
- A bet can be +EV and still lose. EV is a long-run concept, not a per-bet strong value.
- Even small edges (+3 to +5 EV per $100) compound meaningfully across hundreds of decisions.
- The more accurate your probability model, the more reliable your EV estimates become.
How to Estimate True Probability (The Hard Part)
The EV formula is straightforward; the challenge is building honest, accurate probability estimates. This is where most bettors fall short. Sportsbooks employ quantitative analysts and vast data pipelines—your edge, if any, comes from identifying specific market inefficiencies they may underweight or from acting faster on breaking information.
- Line shopping: Compare odds across multiple markets. A line of +130 at one book versus +115 at another represents a real EV difference.
- Closing line value (CLV): If you consistently beat the closing line—the final price before an event starts—you are demonstrating that the market moved toward your number, validating your probability estimate.
- Sharp money indicators: Significant line movement opposite to public betting percentages often signals professional action and a potential market correction.
- Situational and contextual factors: Travel schedules, rest advantages, weather, referee/officiating tendencies, and motivational spots are categories where models sometimes lag public data.
- Sample size discipline: A 10-game hot streak does not validate a model. Meaningful EV assessment requires hundreds of observations.
Common EV Mistakes to Avoid
Understanding the formula is only the beginning. These are the errors that erode edge even when bettors grasp the concept intellectually.
- Overconfidence in probability estimates: Assigning 65% probability to a coin-flip outcome is not analysis—it is bias. Calibration matters.
- Ignoring vig: A −110 line requires a 52.4% win rate just to break even. Many bettors forget this baseline.
- Cherry-picking results: Evaluating EV only on winners is survivorship bias. You must track every bet in your decision set.
- Chasing steam without context: Following line movement blindly copies others' EV estimates without understanding the underlying reasoning.
- Treating EV as certainty: Variance is real. Even a strong +EV bet loses more than a third of the time.
How a Tracked Record Demonstrates EV in Practice
Theory becomes credible when tested against real outcomes over a meaningful sample. GetCloseline.io maintains a publicly tracked, critic-graded pick record across 713 settled bets, currently sitting at 410–303 (57.5% win rate). That win rate is published transparently—not curated to show only the best stretches—because the entire premise of EV-based analysis is that the math should hold at scale, not just in favorable windows.
A 57.5% win rate on spread and total wagers, where the break-even threshold typically sits near 52.4%, is consistent with what sustained +EV selection looks like across a large sample. It does not mean every pick wins—303 losses are right there in the record. It means the process of identifying mathematically favorable spots, applied consistently, produces outcomes that outperform random selection over time. That is expected value working exactly as the formula predicts.
See today's full card