Positive EV Betting: A Sharp Bettor's Guide to +EV
Positive EV betting means only taking prices that beat a sharp fair line. How +EV betting works, from sharp books and devigging to Kelly stakes and CLV.
Positive EV betting (+EV betting) means taking a bet only when the price beats the outcome's fair odds. The fair odds come from sharp sportsbooks, the books that take the largest bets and correct their lines fastest, once their built-in margin is removed. When a soft book pays more than that fair price, the bet has positive expected value: it returns more than it costs on average across many bets.
That is the whole strategy. It is not about picking more winners. It is about paying less than fair for every outcome you back, over a large enough sample for the difference to show. Walk one bet through the full check below, then read on for why each step works.
Please note: Every example in this guide is illustrative, built from prices chosen to show the math, and does not represent any user's results. +EV betting involves variance, depends on a fair price that is only an estimate, and never guarantees a result. Only bet money you can afford to lose.
Try It: One +EV Bet, Start to Finish
Enter a sharp line from up to three books, pick a devig method, then type the price a soft book offers on your side. The check blends the sharp books, strips the vig, prices the edge, and sizes a stake from your bankroll. Toggle a book off, or switch the devig method, and watch the verdict move.
Positive EV bet check
Illustrative only, not a prediction of any result. The sharp line is blended with the same weights, the same 10% cap on each book's vig, and the same devig and EV functions the SmartStake Positive EV tool runs, and the stake uses the tool's Kelly formula. The long-run average assumes the fair price is right, which no one can know for certain, and any 100 real bets land above or below it. Each bet can win or lose, and a stake of $0 means the price carries no edge.
The defaults are a hypothetical NBA moneyline: the Lakers at +110 at Pinnacle, +105 at Circa and +108 at BetOnline, against a soft book offering +125. The blend comes out at +108 / −124, the fair price is about +116, and +125 carries an estimated edge of +4.39%. A quarter Kelly stake on a $1,000 bankroll is $8.78.
What Is EV Betting?
EV betting, short for expected value betting, means comparing a bet's price with its fair odds and placing it only when the price is better. Expected value is the average amount a bet wins or loses per dollar over many repeats. A positive number means the price pays more than the risk is worth. A negative number, which is where almost every bet lands, means the book's margin is working against you.
The formula takes the fair win probability p and the decimal odds o you are offered:
At a fair probability of 46.40% and odds of 2.25 (+125), EV is 0.4640 × 2.25 − 1, or +4.39% per dollar staked. The Expected Value Calculator guide walks the formula in more depth. This guide is about the harder question: where does a trustworthy p come from?
Sharp Books and Soft Books
The betting market has two kinds of sportsbook, and +EV betting lives in the gap between them.
- Sharp books such as Pinnacle and Circa run low margins, take large bets, and move their lines when respected bettors hit them. Every sharp bet is information, so their prices end up close to the true chance of each outcome. The sharp money guide explains how that information moves a line.
- Soft books such as DraftKings, FanDuel and BetMGM are built for recreational bettors. They carry more margin, copy the market more slowly, and manage risk by limiting the accounts that beat them rather than by moving the line.
SmartStake trusts sharp books as the benchmark for an event's fair price, and blends several of them rather than leaning on one. The Positive EV tool weights Pinnacle and Circa at 100, and adds BetOnline and Bookmaker on main lines, FanDuel on player props, and the Sporttrade, Novig, ProphetX and 4Caster exchanges at a weight of 50. It averages the books' odds by weight, then devigs the blend. Any sharp book whose own vig tops 10% is left out.
You can change all of it. The custom devig and sharp books settings let you add or drop books, reweight them, or make one required. Review the evidence and use the sharp books you trust most. The sharpest book also changes by market: our MLB player prop data study found prediction markets like Kalshi and ProphetX leading there.
The Evidence Behind Sharp Lines
Choosing a benchmark is the biggest decision in +EV betting, so here is the public data behind it.
DataGolf analyzed golf matchup odds by using one book's margin-free closing odds to bet against every other book 1. It grouped the bets by the profit those odds expected, then checked what each group actually returned, across 10,972 unique bets priced off Pinnacle and 8,178 priced off DraftKings.
| Expected value group | Pinnacle expected | Pinnacle actual | DraftKings expected | DraftKings actual |
|---|---|---|---|---|
| Below −10% | −12.6% | −13.9% | −13.8% | −9.3% |
| −10% to −8% | −8.9% | −9.3% | −8.9% | −8.5% |
| −8% to −6% | −6.9% | −5.4% | −7.0% | −5.7% |
| −6% to −4% | −5.0% | −5.8% | −5.0% | −5.0% |
| −4% to −2% | −3.1% | −2.5% | −3.0% | −4.0% |
| −2% to 0% | −1.1% | −3.6% | −1.1% | −2.5% |
| 0% to 2% | 0.8% | 0.6% | 0.9% | −5.3% |
| Above 2% | 4.5% | 3.7% | 5.6% | −4.1% |
Read the bottom row. When Pinnacle's odds said a group of bets should return 4.5%, those bets returned 3.7% in the dataset. When DraftKings' odds said a group should return 5.6%, those bets lost 4.1%. A soft book's idea of value was not value at all.
A separate study by Joseph Buchdahl at Football-Data.co.uk tested nearly 88,000 soccer odds pairs and found the same pattern: the profit expected from Pinnacle's closing odds tracked the returns those bets actually produced in the dataset 2.
Past performance in a research dataset does not guarantee future results, and no line is perfect. But the studies point one way. A price that beats a devigged sharp line has been the most reliable public signal of value, which is why SmartStake builds its fair line from sharp books.
Estimating the True Price
A sharp line still carries the book's margin, called the vig or juice. Before you can compare it with anything, you strip that margin out. That step is called devigging.
- Turn each price into an implied probability. Divide 1 by the decimal odds. The blended Lakers price of 2.076 implies 48.17%, and the Warriors at 1.8096 imply 55.26%.
- Measure the vig. The two sides add up to 103.43%, not 100%. The extra 3.43% is the margin.
- Remove it. A devig method shares the extra 3.43% out between the two sides until they total 100%. The power method, the Positive EV tool's default, lands the Lakers at 46.40%, a fair price of about +116.
- Shop the fair price. Any book paying +116 or better on the Lakers is offering a price above fair value.
The method matters more than most guides admit. Power and the simpler multiplicative method disagree by about 0.2 points of probability here, and that is enough to flip a borderline bet:
| Soft book price | EV, power devig | EV, multiplicative devig |
|---|---|---|
| +125 | +4.39% | +4.79% |
| +120 | +2.07% | +2.46% |
| +115 | −0.25% | +0.13% |
| +110 | −2.57% | −2.20% |
At +115, one method calls the bet +EV and the other does not. That is why the check above also shows the worst case EV, the lowest EV across the devig methods, and why thin edges deserve suspicion. The devigging guide compares all four methods, and No Vig Fair Odds explains what a fair price does and does not promise.
Try it on a single book below. Enter both sides of any two-way market to see the fair odds under each method.
Spotting a +EV Line
With a fair price in hand, finding +EV is a comparison. Line up the soft books against it and look for any price that pays more.
| Book (hypothetical) | Lakers price | EV at a 46.40% fair probability |
|---|---|---|
| Soft book A | +125 | +4.39% |
| Soft book B | +120 | +2.07% |
| Soft book C | +110 | −2.57% |
The same bet is +EV at one book and a losing proposition at another. That is why line shopping sits underneath every +EV strategy, and why the number of prices you can reach matters more than the number of accounts you hold. How Many Sportsbooks Do You Need? runs that math.
Notice what the fair price does not tell you. It says nothing about who will win tonight. The Lakers lose this game about 54% of the time at any price. +EV is a statement about the price, not the pick.
Sizing the Bet
An edge only turns into a long-run result if you survive the swings in between. Stake size is how you control that.
The Kelly criterion sizes a bet from its edge and its odds: bet the fraction of your bankroll that maximizes long-run growth. For the Lakers at +125 with a 46.40% fair probability, full Kelly is 3.51% of the bankroll. Most bettors use a fraction of it, because the fair probability is only an estimate and full Kelly swings hard when that estimate is off.
The Positive EV tool suggests a quarter Kelly stake by default. On a $1,000 bankroll that is $8.78, under 1% of the bankroll on a bet with a 4% edge. Set the multiplier in the tool's bankroll and Kelly settings. The Kelly Criterion Calculator guide and Bankroll Management cover staking in full.
Modeling Expected Returns
Here is the question everyone asks next: what do many bets like this add up to?
Multiply the stake by the edge and the number of bets. 100 quarter Kelly bets of $8.78 at +4.39% have a long-run average of about $38.57, on $878 staked. That is an average, not a forecast. With a win rate under 50%, runs of 5 or 6 losses in a row are routine, and 100 bets is a small sample. Any real 100 bets land above or below the average, sometimes far below it.
That gap between a real edge and a visible result is why +EV bettors do not judge themselves on profit and loss.
Measuring Whether It Works
Profit and loss is too noisy to confirm an edge for thousands of bets. Closing line value (CLV) is faster. It compares the price you took with the final sharp price before the game starts. If you keep beating the closing line, your prices were better than the market's final, most informed estimate, which is the same claim +EV betting makes.
Drag the sliders to see how many bets it takes before an average CLV stands out from the noise.
Is your edge real, or a lucky run?
Set your average closing line value per bet, how much it swings bet to bet, and how many bets you have tracked. The same math the SmartStake Performance dashboard runs then scores how confident you can be that the edge is skill, not variance.
Illustrative only. Confidence is a one-sided normal test on your mean CLV, the same function behind the Performance dashboard edge significance readout. Closing line value is a leading indicator, not a promise. A positive read means your prices look sharp over the sample, not that any single bet wins.
At an average CLV of +2% and a 9 point swing from bet to bet, both assumed rather than measured, the test clears 95% confidence at about 55 bets. Profit and loss alone can take thousands.
The closing line value guide explains the idea, How Closing Line Value Is Calculated covers the four ways to measure it, and How to Track Sports Bets shows the log columns that make it possible. The SmartStake Bet Tracker scores CLV on your settled bets automatically.
Finding +EV in Practice
Price gaps between sharp and soft books appear every day. They also close fast. When a soft book hangs a price well above fair, sharp bettors hit it, the book sees the money, and the line moves. Many edges last minutes, some only seconds.
That makes the manual version of this guide hard to run at scale. Comparing dozens of books across hundreds of markets by hand means most edges are gone by the time you find them.
The SmartStake Positive EV tool automates the whole check above. It compares 70+ sportsbooks, exchanges and DFS sites against a blended, devigged sharp line, and lists every price that clears it with its EV and fair odds. Take a bet and it suggests a Kelly stake from your bankroll. Pregame lists refresh on demand or automatically every 10, 30 or 60 seconds, and a live mode tracks in-play prices. Finding Positive EV Bets documents the workflow, and How to Use the SmartStake Positive EV Tool walks through it screen by screen.
A few habits separate a sustainable approach from a short one:
- Skip the thinnest edges. An edge that only one devig method sees is noise more often than value.
- Distrust the biggest ones. An EV above 10% usually means a stale line, a mismatched market or a book about to void the bet.
- Plan for limits. Soft books cut stakes for accounts that keep beating the close. Expect it rather than hoping to avoid it.
- Log everything. Without CLV on every bet, you cannot tell a real edge from a lucky run.
Results vary and are never guaranteed. +EV betting is a long-term, high-volume approach that takes discipline and careful bankroll management.
See the live +EV feed with SmartStake
Frequently Asked Questions
What is EV betting?
EV betting, short for expected value betting, means placing a bet only when the price pays more than the outcome's fair odds. The fair odds come from sharp books with the vig removed. A bet whose price beats them has positive expected value (+EV): it returns more than it costs on average over many bets, though any single bet can still lose.
Does positive EV betting guarantee a profit?
No. Positive EV betting describes the average of many bets, not the result of any one of them, and a real edge of 2% to 5% can still sit behind a losing month. The fair price is an estimate, prices move, and sportsbooks limit accounts that keep beating them. Treat +EV as a way to find value, and only bet money you can afford to lose.
Which sportsbook sets the fairest line?
Pinnacle and Circa are the benchmark sharp books for most main markets, because they take large bets and move their lines on sharp action. SmartStake blends them at full weight with several exchanges and sharp books at half weight, and you can change the books and weights. Which book is sharpest varies by sport and market.
How much EV makes a good bet?
There is no single cut-off, because the fair price is an estimate. Many bettors ignore edges under about 1% to 2%, where a small change in the devig method can flip the verdict, and treat edges above 10% with suspicion, since they often mean a stale or mistaken line. Checking the worst case EV across devig methods is a simple guard.
Will sportsbooks limit you for positive EV betting?
Often, yes. Soft sportsbooks profile accounts that keep beating the closing line and cut their maximum stakes, sometimes within weeks. Spreading action across books and mixing bet types can slow that down, but no approach prevents it, so plan for limits from the start.
Further Reading
- Best EV Betting Software compares the +EV tools on price and coverage.
- Best EV Betting Tools for Canada ranks them for Ontario and Alberta bettors.
- How to Build a Sports Betting Model covers finding a fair price from your own model instead of the market.
- The free Devigging Calculator and Expected Value Calculator run the steps in this guide on any line.
References
Footnotes
-
DataGolf. (2020). How Sharp Are Sports Bookmakers? An analysis of closing line value in golf betting markets. https://datagolf.com/how-sharp-are-bookmakers ↩
-
Buchdahl, J. (2016). Pinnacle's Closing Line and Your Betting Yield. Football-Data.co.uk. https://www.football-data.co.uk/blog/pinnacle_efficiency.php ↩
Get the next guide in your inbox
Practical betting guides, new tools, and product updates from the SmartStake team. Unsubscribe in one click, any time.
By providing your email, you are consenting to receive communications from SmartStake. Visit our Privacy Policy for more info
