Learn the +EV formula, break-even probability, and how to measure edge before placing a sports bet.
New to this? Start with:How to Use the +EV Calculator
You found Dodgers +140 on FanDuel, and your model says they win 47% of the time. Is that worth a bet? The answer lives in one number: expected value (EV).
EV tells you whether a bet is worth taking over the long run. It does not predict tonight's result — it answers a sharper question: if you could place this exact bet thousands of times, would you come out ahead or behind? Every profitable sports bettor builds their process around this concept. This guide breaks down the formula, walks through three worked examples, and gives you a practical workflow you can use before every bet.
Expected value is the weighted average of all possible outcomes. You multiply each outcome by its probability, then add the results together. The number you get is the amount you can expect to gain or lose per bet, on average, over a large sample.
Think of a simple coin flip. Someone offers you a game: heads you win $120, tails you lose $100. The coin is fair, so each side has a 50% chance. Your expected value on every flip is:
You expect to make $10 per flip on average. That is a game you want to play. Now flip the payout: heads you win $80, tails you lose $100. Same fair coin.
Now you expect to lose $10 per flip. That is a game you want to avoid.
Sports betting works the same way. The odds set the payout. Your job is to estimate the probability. When your probability estimate says the payout is generous enough, the bet has positive expected value.
Why EV matters more than win rate: A bettor who wins 55% of -110 bets is profitable. A bettor who wins 60% of +300 bets is extremely profitable. A bettor who wins 70% of -500 bets may be barely breaking even. Win rate alone tells you nothing without the payout context. EV combines both into one number.
The core formula for expected value on a single bet is:
EV = (P_win x Profit) - (P_lose x Stake)
Where:
You can also write the formula as:
EV = (P_win x Decimal Odds x Stake) - Stake
This version is convenient when you already have decimal odds and want to skip the intermediate profit step. Both forms produce the same result.
To compare bets of different sizes, convert EV to a percentage of your stake:
EV% = (EV / Stake) x 100
EV% is your expected return on investment for that bet. A bet with EV% of +5% means you expect to earn 5 cents for every dollar wagered, on average, over time.
Every set of odds has a built-in break-even point: the win rate at which you neither make nor lose money over time.
Break-Even % = (1 / Decimal Odds) x 100
If your estimated win probability is above the break-even percentage, the bet has positive expected value. If it is below, the bet has negative expected value.
| American Odds | Decimal Odds | Break-Even % |
|---|---|---|
| -300 | 1.33 | 75.0% |
| -200 | 1.50 | 66.7% |
| -150 | 1.67 | 60.0% |
| -110 | 1.91 | 52.4% |
| +100 | 2.00 | 50.0% |
| +150 | 2.50 | 40.0% |
| +200 | 3.00 | 33.3% |
| +300 | 4.00 | 25.0% |
| 1.50 (decimal) | 1.50 | 66.7% |
| 3.00 (decimal) | 3.00 | 33.3% |
What "above break-even" means in practice: If a line is priced at +150 (decimal 2.50), the break-even probability is 40%. If your model says this team wins 45% of the time, you have a 5-percentage-point edge. That gap between your estimated probability and the break-even probability is your edge, and it is the source of your expected profit.
Use the Odds Converter to quickly switch between American, decimal, and fractional formats, or the No-Vig Calculator to strip the bookmaker margin from a line and see the fair odds.
Scenario: You want to bet on the Los Angeles Dodgers moneyline. The sportsbook offers +140 (decimal 2.40). Your model projects the Dodgers to win 47% of the time.
Step 1: Find break-even probability.
Step 2: Compare to your projection.
Step 3: Calculate EV.
Step 4: Calculate EV%.
This is a strong +EV bet. You expect to earn $12.80 for every $100 wagered at these odds and this probability.
Scenario: A sportsbook offers the Kansas City Chiefs at -180 (decimal 1.56). Your model projects the Chiefs to win 60% of the time.
Step 1: Find break-even probability.
Step 2: Compare to your projection.
Step 3: Calculate EV to confirm.
Step 4: Calculate EV%.
This is a negative EV bet. Even though the Chiefs are likely to win, the price does not offer enough payout to justify the risk. You should pass.
Scenario: A sportsbook offers the Milwaukee Bucks at -110 (decimal 1.91). Your model projects the Bucks to win 54% of the time.
Step 1: Find break-even probability.
Step 2: Compare to your projection.
Step 3: Calculate EV.
Step 4: Calculate EV%.
Technically positive, but the edge is thin. Whether you take this bet depends on how confident you are in your 54% estimate and your personal EV% threshold. A 1-2 percentage point error in your model flips this bet to negative EV. Many sharp bettors set a minimum threshold of 2-3% EV to account for model uncertainty. At 3.14%, this is right at the boundary. If your model has been well-calibrated over hundreds of predictions, it may be worth taking. If you are less confident in your projections, passing is reasonable.
EV% is essentially your expected ROI per bet. If you consistently place bets with +5% EV, your long-term ROI should converge toward 5% of total volume wagered.
However, "long-term" matters. Over a small sample of 20 or 50 bets, anything can happen. Variance is real. A bettor with a genuine 5% edge can easily have a losing month. The math works in your favor only when applied across hundreds of bets.
How to think about variance: Imagine you flip a slightly weighted coin that lands heads 55% of the time. In 10 flips, you might get only 3 heads. In 100 flips, you are very likely to see somewhere between 45 and 65 heads. In 1,000 flips, you will almost certainly see close to 550 heads. The larger your sample, the more reliably your actual results match your expected results.
This is why bankroll management matters. You need enough capital to survive the inevitable losing streaks so you are still in the game when the math catches up. A common guideline is to risk no more than 1-3% of your bankroll on any single bet.
Use this six-step process before every bet:
Find the market odds. Check the sportsbook price and note the decimal odds. Use the Odds Converter if the line is in American or fractional format.
Convert to break-even probability. Divide 1 by the decimal odds and multiply by 100. This is the minimum win rate the bet needs to break even.
Estimate your true probability. Use your model, historical data, or sharp consensus lines to estimate the actual win probability. The No-Vig Calculator can help you derive fair probabilities from the market itself by removing the bookmaker margin.
Calculate EV and EV%. Plug your numbers into the formula, or use the +EV Calculator for an instant result. Check that EV% is positive and above your minimum threshold.
Compare edge size to your threshold. Most sharp bettors require at least 2-3% EV before placing a bet. Smaller edges carry more risk of being erased by model error.
Track results to validate your model. Log every bet, your projected probability, the actual odds, and the outcome. Over time, compare your predicted win rates to your actual win rates. If your 55% projections are winning 48% of the time, your model needs adjustment.
Not every positive EV bet is worth taking. Consider passing when:
Tip: The biggest mistake is not running the formula wrong — it is using bad probability inputs. Spend 80% of your time improving your probability estimates and 20% on the EV math. The formula is easy; the projection is the hard part.
Open the +EV Calculator and plug in tonight's lines. Enter your own probability estimate and see whether the EV is positive, negative, or borderline. Then compare prices across 2-3 sportsbooks — even a small odds improvement can flip a marginal bet into a clear +EV spot.
Expected value is (win probability x profit if win) - (lose probability x stake). A positive result indicates a +EV bet.
Compare your estimated probability to the break-even probability from the market odds. If your estimate is higher, the bet can be +EV.
EV% is your expected value divided by your stake, multiplied by 100. It tells you the percentage return you expect on every dollar wagered over the long run.
Yes. A single +EV bet can lose. Expected value describes the average outcome over many repetitions, not the result of one bet. Variance means you will have losing streaks even with a genuine edge.
Many sharp bettors look for at least 2-3% EV before placing a bet. Smaller edges can be eaten up by variance and model uncertainty, so a higher threshold provides a margin of safety.
Break-even probability equals 1 divided by the decimal odds, multiplied by 100. It is the minimum win rate you need to avoid losing money at those odds over time.
There is no exact number, but most analysts suggest at least 500 to 1,000 bets at a consistent edge before you can reliably distinguish skill from luck. The smaller your edge, the more bets you need.