Positive EV (+EV)

In short: A bet where the offered odds imply a lower probability than the true probability of the outcome — meaning the expected return over many repetitions is positive. Long-term profitability in sports betting comes from consistently placing +EV bets, regardless of whether any individual bet wins or loses.

A +EV bet exists whenever you can buy a position at a price below its fair value. The concept is simple. The hard part is sticking with the math while you’re a few dozen bets in and watching your account swing through a losing streak that has nothing to do with whether your edge is real.

How a +EV opportunity actually looks

Consider a real workflow. The Yankees are playing the Red Sox tonight. DraftKings is offering Yankees moneyline at +118. Pinnacle, the sharp reference book most serious bettors use as their fair-price benchmark, is offering Yankees +112 / Red Sox -125.

The first step is to remove the vig from Pinnacle’s prices to recover the implied true probabilities:

• Yankees +112 → 47.2% vig-included implied probability
• Red Sox -125 → 55.6% vig-included implied probability
• Combined: 102.8% (the 2.8% excess is Pinnacle’s vig)
• Devigged: Yankees 45.9%, Red Sox 54.1%

Now compare the DraftKings price to the devigged Pinnacle estimate:

• DraftKings Yankees +118 implies 45.9% probability
• Pinnacle devigged says Yankees are actually 45.9% likely to win

Identical. Zero edge. This would be a fair price, not a +EV bet.

But if DraftKings were instead offering +135 on the Yankees, that’s a different story:

• +135 implies 42.6% probability
• Pinnacle devigged still says 45.9%
• The gap is 3.3 percentage points. That’s your edge.

In expected-value terms:

EV = (P_win × profit) - (P_loss × stake)
EV = (0.459 × $135) - (0.541 × $100)
EV = $61.97 - $54.10
EV = +$7.87 per $100 staked

That’s a 7.87% edge. Bet it whenever you find it.

Why sharp books are the reference

The choice of reference matters more than the formula. The math is the easy part. Picking a fair-price source that actually reflects fair value is what separates real +EV bettors from people who think they’re +EV bettors.

A handful of books accept enough sharp action that their lines reflect real fair value rather than retail vig padding:

• Pinnacle accepts five-figure first bets without limiting and adjusts lines on sharp action. The international standard reference, but unavailable to US bettors directly.
• Circa Sports posts NFL openers earlier than competitors and welcomes sharp action. Available in Nevada and a handful of other states.
• Novig is a peer-to-peer exchange with no vig. Prices come from supply and demand among users, not a bookmaker’s model.

These are the books whose lines you want to devig and compare against. Retail books like DraftKings, FanDuel, and BetMGM are not appropriate references. Their lines include 4-6% vig and are shaped by where the public is betting, not where the price should be.

A common mistake: averaging the lines of every available book to estimate “the market consensus” and using that as the fair reference. This pollutes the signal with retail vig and public shading. One sharp reference, devigged cleanly, beats a 30-book average.

What +EV is not

Three concepts that get conflated:

+EV vs. closing line value. CLV measures whether you got a better price than the eventual closing line. +EV measures whether your bet beat the sharp consensus at the moment you placed it. They’re related but not the same. CLV is the better short-term skill signal because it’s less noisy than win/loss results, but +EV is the bet you’re actually making.

+EV vs. arbitrage. Arbitrage guarantees profit on every outcome by betting both sides at different books. +EV is one-sided and probabilistic. You expect to profit on average, but any individual bet can lose. Arbs have effectively zero variance per bet; +EV bets have substantial variance.

+EV vs. “value betting.” “Value bet” is sometimes used interchangeably with +EV, but the term is loose enough that it gets applied to any bet someone subjectively likes. True +EV requires a quantitative comparison to a fair reference. If you can’t write down the no-vig price you’re comparing against, you don’t actually know whether the bet is +EV.

The hard parts nobody mentions

Most +EV explainers stop at the math. The math is the easy part. The harder parts:

The headline edge is smaller after friction. A theoretical 7% EV bet in a perfect world becomes closer to 4-5% real EV after accounting for line moves between you spotting the opportunity and placing the bet, errors in your devigging assumptions, bonus bet restrictions and rollover requirements, withdrawal fees, taxes, and the time cost of running a multi-account operation. Most of these are bearable, but they compound.

Variance is real even when you’re right. The size of your edge directly determines how long it takes for win/loss results to separate from noise. A 7% blended edge, typical for users running a scanner against retail books, starts showing a reliable skill signal within 100-300 bets. By the upper end of that range, you can be reasonably confident you’re profitable rather than running hot. Smaller edges (1-3%) take dramatically longer to confirm, sometimes thousands of bets. The math eventually delivers regardless. The question is how patient you can be in the meantime. See the variance entry for sample-size detail across different edge sizes.

CLV is the better short-term feedback signal. Win/loss results take hundreds of bets to converge. Closing line value converges much faster. Within dozens of bets, you can tell whether you’re consistently getting prices that beat the closing line. If you are, the math will work out eventually, even if you’re running cold this month.

What “consistently +EV” looks like as a habit

The serious +EV bettor’s workflow doesn’t look like picking games. It looks like:

1. Run a scanner (or manually compare lines across multiple books) to find prices that diverge from a sharp reference
2. Devig the sharp reference to get a fair-probability estimate
3. Place the retail-book bet if the edge is large enough to clear the variance threshold (typically >2-3%) and the bet fits within bankroll and Kelly sizing
4. Log the bet, the price taken, and the closing line for CLV tracking
5. Move on. Don’t agonize over the outcome. Outcomes don’t tell you whether the decision was right.

The bets themselves are often boring: -110 spreads, totals, alt lines on player props. The skill is in the process discipline, not picking winners.

A practical note on limits

The books offering +EV opportunities are doing so because their pricing isn’t perfectly sharp. Many retail books track signs of sharp betting and will eventually reduce your maximum bet sizes on an account they’ve flagged.

This is operational friction, not a strategic problem. The +EV game is structured around having accounts at many books. Ten to 20 is typical for serious bettors. If one book limits you, you have plenty of others. Limited accounts also often get unrestricted again after a quiet period, so even the worst case is rarely permanent.

For specifics on what triggers limits and how to extend account lifespan, see account limiting.

For more on finding and evaluating +EV opportunities in practice, see +EV Betting Explained.