Expected value is the single most important concept in sports betting. Every professional bettor, every sharp, every syndicate operation organizes its entire approach around one question: does this bet have positive expected value? If the answer is yes over a large enough sample, money flows toward you. If the answer is no, the house wins regardless of how good your gut feels in the short term. This guide breaks down the EV formula, shows you how to calculate it on real bets, explains how to find +EV opportunities, and gives you the tools to verify your edge over time.
Expected value measures the average outcome of a repeated decision over many trials. In betting, the formula is:
Every variable matters. The probability of winning is your honest, calibrated estimate — not the odds you were given, not wishful thinking. The net profit is what you clear if you win (your total return minus your stake). The stake is what you lose if the bet fails.
Let's build a concrete example. You bet $110 at -110 odds on a spread. The sportsbook returns $210 if you win ($100 profit + your $110 back). You estimate this team covers 55% of the time based on your research.
Over 100 bets at that EV, you expect $550 in profit. That's not guaranteed — variance will scatter your actual results around that expectation — but it's the long-run mathematical truth of the bet.
Now let's see what negative EV looks like. Same bet, same -110 odds, but this time your true win probability is 50% (exactly the sportsbook's implied probability).
Over 100 bets, expected loss is $500. The vig is eating you alive. This is what happens when you bet without an edge — you're paying the house 4.76% on every -110 bet. Over thousands of bets, that compounding loss is catastrophic for recreational bettors who don't track it.
The EV formula also scales proportionally. A $550 bet at -110 with 55% true probability: EV = (0.55 × $500) − (0.45 × $550) = $275 − $247.50 = +$27.50 per bet. Higher stakes amplify both upside and downside.
Before you can calculate EV, you need to know what probability the sportsbook is pricing in. That's the implied probability — the win probability that would make the bet exactly break-even (zero EV).
American negative odds (favorites): Implied probability = |odds| ÷ (|odds| + 100)
American positive odds (underdogs): Implied probability = 100 ÷ (odds + 100)
Decimal odds: Implied probability = 1 ÷ decimal odds. At 2.50, implied = 1 ÷ 2.50 = 40%. At 1.909, implied = 1 ÷ 1.909 = 52.38%.
Here's the critical piece: on a standard -110/-110 spread market, each side has 52.38% implied probability. Add them up: 52.38 + 52.38 = 104.76%. The total exceeds 100% by 4.76 percentage points — that excess is the vig (also called juice or margin). The sportsbook collects that margin regardless of outcome. On every single market, the implied probabilities always sum above 100%, and the excess is always the house's take.
A -120/+100 market: -120 implies 54.55%, +100 implies 50.00%, total = 104.55% — vig of 4.55%. Some books run tighter markets (3–4% vig), some looser (5–7%). Knowing the vig on a market helps you understand how much edge you need just to break even.
The core requirement for a +EV bet is simple to state and hard to execute: your honest probability estimate for the outcome must exceed the implied probability in the odds. That gap — if it exists and is real — is your edge.
If a team is priced at +150 (implied 40%) and your model gives them a 46% true chance of winning, you have a 6-percentage-point edge. Let's see the EV:
That's a 15% edge — exceptional. Real edges are usually far smaller: 2–5% is good, 5–10% is elite. Most recreational bettors overestimate their edges by a factor of 2–3x, which is why bankroll management matters even when you think you've found a great bet.
The hardest part of finding +EV isn't the math — it's building calibrated probability estimates. Ways serious bettors approach this:
A useful rule of thumb: if you cannot articulate a specific reason why your probability estimate differs from what the market is implying, you probably don't have an edge. "I feel good about this team" isn't a probability estimate — it's noise.
Short-term betting results are dominated by variance. You can win 60% of your bets over 50 games and still be a long-term losing bettor who got lucky. You can lose 45% of your bets over the same stretch and be a long-term winner who ran cold. Results alone don't tell you whether your process is +EV until you have thousands of bets in the sample.
The closing line solves this problem. The closing line is the final odds posted by a sportsbook just before a game starts. It is set by sharp money — professional bettors with sophisticated models who have bet heavily enough to move the line. The consensus closing line at major books is considered the most accurate probability estimate of the game outcome available.
Closing Line Value (CLV) measures whether the odds you bet were better or worse than the closing odds:
Professional bettors track CLV religiously. If your average CLV is positive across 200+ bets, it strongly suggests your process identifies value. The closing line is a proxy for "true probability" — consistently beating it means you're finding +EV before the market corrects.
Important caveats: CLV is most meaningful at books with sharp action (Pinnacle, Circa, Bookmaker). Soft books sometimes move lines for reasons unrelated to true probability (balanced action, risk management). Track CLV at multiple books for the cleanest signal. Also note: even CLV tracking needs a sample of at least 200–300 bets before the noise resolves into a meaningful signal.
The practical takeaway is this: track every bet with the opening line, your bet price, and the closing price. After 300 bets you'll have a far more honest picture of your edge than your win-loss record gives you.
Understanding who moves lines — and why — is essential for interpreting market signals and finding value before it disappears.
Sharp money comes from professional bettors and syndicates who use quantitative models, large data sets, and significant bankrolls. Their bets move lines. Sharps are not infallible — they win roughly 55–58% of their bets against the spread, not 80% — but they are far more calibrated than the public and their sustained edge is real.
Square money (public betting) comes from recreational bettors acting on brand recognition, media narratives, and recent visible performance. Public money tends to flow toward popular teams, heavy favorites, overs, and high-profile matchups regardless of value. Books often shade lines to capitalize on public tendencies.
Key line movement concepts:
Tools to track this: Action Network's public betting percentages, ESPN BET's percentage tool, and Sharp Action data from OddsChecker all provide real-time looks at where money and tickets are flowing. None of these signals guarantee a win — they tell you where informed money is going, which is useful information when building your own probability estimate.
This section is arguably more important than the EV formula itself, because misunderstanding variance destroys bankrolls even when bettors have genuine +EV.
Positive EV means you will profit in the long run — not on every bet, not in every session, not even in every 50-bet stretch. Variance is the natural statistical spread of outcomes around your expected value. In betting, variance is enormous because individual game outcomes are binary (win or lose) and even the best bettors win only 55–58% of their -110 bets.
Consider a bettor with a genuine +5% edge on -110 lines (55% win rate). Their EV per $110 bet is +$5.50. Now consider the losing streaks that are mathematically normal for this bettor:
This is why bankroll management is not optional for +EV bettors — it's the mechanism that keeps you in action long enough for EV to manifest. Two primary approaches:
Practical bankroll rule: never bet more than 3–5% of your total bankroll on any single game, even if you think you have a huge edge. Your edge estimate is always noisier than it feels. The goal is staying in action, not maximizing any single bet.
Every strategy discussed so far requires you to be better than the market at estimating game probabilities — which is genuinely hard. Line shopping is different. It's a source of positive EV that requires zero predictive skill. You simply take the best available price.
Different sportsbooks post slightly different odds on the same game. These differences arise from different risk models, different customer bases, different liability positions, and timing differences in adjusting to sharp action. A well-organized bettor with accounts at 4–5 books will consistently find better prices than someone locked into a single book.
A concrete example: you want to bet Team A on the moneyline.
You always take Caesars at +155. Over 1,000 bets of $100 each where the true probability is 40%, the difference between always taking +155 vs. always taking +148 is:
That $2,800 difference comes purely from taking 7 extra cents on the odds — no skill required, just discipline. Industry research suggests maintaining 4–5 book accounts improves your effective line by 0.5–1.5 points per bet on average, worth roughly 0.5–2% ROI improvement over time. For a bettor doing $50,000 in annual volume, that's $250–$1,000 in free money.
Setting up accounts at DraftKings, FanDuel, BetMGM, Caesars, and ESPN BET takes approximately 30 minutes total. Each typically offers signup bonuses that add further value. After setup, checking multiple books before every bet becomes a 30-second habit that compounds significantly over years of betting. Treat it as the floor, not the ceiling, of your edge-finding process.
Other line shopping tools: OddsJam, The Action Network's odds screen, and OddsChecker all display real-time lines across books in a single view, making comparison instantaneous. No serious bettor skips this step.