Positive EV betting (+EV betting) is a method, not a pick: you estimate the fair price of an outcome, usually from Pinnacle's odds with the margin removed, and you only bet when a bookmaker offers longer odds than that fair price. Each bet can still lose; the method is judged over hundreds of bets, by closing line value rather than by results.
A single value bet is one price that is too long. Positive EV betting is the routine built on that idea: where the fair price comes from, which edges you take, how much you stake, how long before results mean anything, and how you check the edge was real.
Definition. Positive expected value (+EV) means the average return of a bet, if it could be repeated many times at the same price and probability, is above zero. EV per unit staked = p × odds − 1, where p is the true probability and odds the decimal price. Positive EV betting is the practice of placing only bets where that number is positive, measured against a fair price you can check later.
What does +EV mean in sports betting?
Every bet has an expected value. It is the result you would average if you could place the same bet, at the same price and the same true probability, thousands of times.
- At a true probability of 50% and odds of 2.10: 0.50 × 2.10 − 1 = +5%. On average, each unit staked returns 1.05.
- At 50% and odds of 1.95: 0.50 × 1.95 − 1 = −2.5%.
Most bets at most bookmakers are slightly −EV, because the price includes the bookmaker's margin. Pinnacle's help centre puts its own average margin at about 2%, against an industry average of around 6% (Pinnacle Help Center, "What is a betting margin?", read 28/10/2026). A +EV bet is the exception: a price that is longer than fair, usually because the bookmaker has not updated yet or prices the market loosely. How the margin is built into odds is explained in overround, vig and margin.
EV is a property of the decision, not of the result. A +5% bet that loses was still a +5% bet. A −8% bet that wins was still a bad price.
How does positive EV betting work? The 4-step method
The method has four steps. The first is where nearly all the difficulty lies.
Step 1: Get a fair price you can trust
You need an estimate of the true probability. There are two ways:
- Your own model. Ratings, expected goals, injuries, schedule. It works only if your model is better than the market, and you find that out after hundreds of bets.
- A sharp market. Take the price of a bookmaker that accepts large bets from winning customers and moves on their money, and remove its margin. Pinnacle is the usual choice. Why sharp books make a better reference than recreational ones is covered in sharp vs soft bookmakers; the full step-by-step for using Pinnacle as a sharp reference price includes when not to trust it.
Removing the margin is called devigging. Pinnacle Terminal and the free devig calculator use the power method by default; the devig methods guide compares power, multiplicative, additive and Shin with worked numbers.
Step 2: Compare with the price you can actually get
Edge = your odds ÷ fair odds − 1. It gives the same number as p × odds − 1. Compare like with like: same market, same line, same period. Handicap −1.5 and −2.5 are different bets. A price that the bookmaker will not accept at your stake is not a price. If several of your books offer the market, take the longest one; line shopping shows how much that is worth.
The EV calculator does steps 1 and 2 in one go: enter Pinnacle's two or three prices and your bookmaker's price. If you would rather have software scan bookmakers for you, scanners are compared in best value betting software, and one newer +EV scanner is set against a Pinnacle terminal in EV Kings alternative.
Step 3: Decide which edges to take
Not every positive number is worth a bet. Filters that +EV bettors commonly apply:
- A minimum edge. Small edges are the most sensitive to errors in the fair price. A 0.5% edge disappears if the devig method or a stale Pinnacle price is slightly off.
- A reference with real limits. A fair price from a market that accepts only small bets is a weaker estimate than one from a main line. See low-limit odds moves.
- Fresh prices. The edge exists only until your bookmaker updates. Most practical +EV bets appear in the minutes after Pinnacle moves; this is why dropping odds matter.
- Bets you can place. Account limits at soft bookmakers decide how much of an edge you can actually take. See why bookmakers limit winners.
Step 4: Stake small and record everything
Stake a small, consistent share of your bankroll. The Kelly criterion gives the theoretical maximum stake for a given edge; most bettors use a fraction of it. Try the Kelly calculator. Then record the odds you took and Pinnacle's fair closing price for every bet, so you can check your edge against the closing line (see below).
A real example: +EV prices after a Pinnacle drop
From Pinnacle Terminal on 28 October 2026 (Madrid time, CET). ITF Women Taby, first round, Lois Bernstein v Laetitia Pulchartova, start 13:00. Market: handicap, Bernstein −1.5.
- Pinnacle's price fell from 3.52 to 2.88, and the move was labelled Sharp drop (down 20.3% from its 09:04 high).
- Pinnacle's fair price at the alert, after removing the margin: 3.138.
- Fair probability: 1 ÷ 3.138 = 31.9%.
- The limit on the line was $225.
At that moment, any price above 3.138 on Bernstein −1.5 was +EV against Pinnacle's estimate. The bookmaker prices below are illustrative, not quotes from a real bookmaker:
| Price at your bookmaker | EV vs fair 3.138 | +EV? |
|---|---|---|
| 3.40 | +8.3% | Yes |
| 3.30 | +5.2% | Yes |
| 3.20 | +2.0% | Yes, small |
| 3.138 | 0% | Fair |
| 3.10 | −1.2% | No |
| 2.88 (Pinnacle's own new price) | −8.2% | No |
Two cautions from the same example. First, a $225 limit is thin, so little money was needed to move this line; the fair price is a weaker estimate than it would be on a main market. Second, at a 31.9% probability, this bet loses about two times in three even when the price is +5%. Full Kelly at 3.30 would be about 2.2% of bankroll, and a thin market is a reason to bet a fraction of that.
How many bets until +EV shows in results?
More than most people expect. Take a simple case: every bet at odds of 2.00 with a true probability of 51%, so every bet is exactly +2% EV. With a 1-unit stake, the expected result and the spread of possible results (normal approximation, one standard deviation) are:
| Bets | Expected result | One standard deviation | Chance of still being behind |
|---|---|---|---|
| 100 | +2 units | ±10 units | about 42% |
| 500 | +10 units | ±22 units | about 33% |
| 1,000 | +20 units | ±32 units | about 26% |
| 2,500 | +50 units | ±50 units | about 16% |
| 5,000 | +100 units | ±71 units | about 8% |
These figures are arithmetic, not a forecast: they assume every bet really is +2%, which in practice you never know for sure. The point is the shape. After 1,000 genuinely +EV bets, roughly one run in four is still behind. Results over a few weeks tell you very little in either direction. Test your own numbers in the betting variance simulator.
How do you know your +EV bets were real?
Use the closing line. Pinnacle's price just before the start reflects everything the market learned. If the odds you took are consistently longer than Pinnacle's fair closing price, your bets were +EV by the best available measure, whatever the results. If they are consistently shorter, the "edge" came from a bad fair price, slow execution or a market you misread.
This is closing line value (CLV). It needs far fewer bets than results do, because it does not depend on who won. The closing line value guide explains raw versus no-vig CLV, and the CLV calculator works it out for one bet or a batch.
Positive EV betting vs arbitrage and matched betting
| Method | What you need | Risk on each bet | Main constraint |
|---|---|---|---|
| Positive EV betting | A fair price (usually de-vigged Pinnacle) and a longer price elsewhere | Each bet can lose; the edge is an average | Variance and soft-book limits |
| Arbitrage | Prices at two or more books that together cover every outcome above 100% return | Low if every leg is accepted at the quoted price | Very short windows, stake limits, voided legs |
| Matched betting | Bookmaker promotions plus an exchange lay | Low on paper; depends on executing both sides | Only as long as promotions last |
Arbitrage is covered by the arbitrage calculator. All three look for mispriced odds; only +EV betting accepts that individual bets lose.
Common mistakes
- Using the raw Pinnacle price as "fair". It includes the margin. Devig first, or you will underestimate small edges and overestimate others.
- Trusting thin markets as much as main lines. A fair price on a small limit can be wrong by more than your edge.
- Chasing the biggest edge on the screen. Very large edges are more often stale prices, wrong lines or suspended markets than real opportunities.
- Staking up after a winning week. EV does not change because of recent results.
- Judging the method by results. Judge it by CLV over a few hundred bets. Results come later, and only on average.
How Pinnacle Terminal handles this
Pinnacle Terminal shows Pinnacle's live price and its fair price (NVP) side by side on every line, with power as the default devig method and multiplicative, additive, Shin and a worst case available. Every pre-match move is graded 0–10 and labelled (Big move when the odds drop and Pinnacle closes the market, Sharp drop when they fall about 20% or more from their recent high, Limit-backed when the odds drop and Pinnacle raises the limit, plus Downtrend, Rebound and Standard), so you can see which moves to compare against your bookmaker first. Live in-play drops are shown as they happen and filtered by match context, without a grade. The tracker records the odds you took and calculates CLV against the fair closing price automatically. The terminal does not accept bets and is not affiliated with Pinnacle. See every graded move on the Pinnacle dropping odds page.
FAQ
What is positive EV betting in simple terms?
Only betting when the odds you are offered are longer than the fair odds of the outcome. At a fair price of 2.00, taking 2.10 is +EV; taking 1.95 is not.
How do you calculate EV in betting?
Multiply the true probability by the decimal odds and subtract 1. At 51% and odds of 2.00: 0.51 × 2.00 − 1 = +2% per unit staked.
Is +EV betting the same as value betting?
Yes. A value bet is a +EV bet. "Positive EV betting" usually refers to doing it systematically, with a fair-price source, filters, staking rules and CLV tracking.
Can you lose with positive EV betting?
Yes. Individual bets lose often, and even a genuine +2% edge can be behind after 1,000 bets. EV describes the average of many bets, and only if your fair price is right.
Where do +EV bets usually come from?
Mostly from soft bookmakers that have not yet copied a sharp price change, and from markets they price loosely. The window usually closes once the bookmaker updates.
Try it
See Pinnacle's fair price next to every line, with every pre-match move graded 0–10 and live in-play drops as they happen: start the Pinnacle Terminal 7-day free trial. Nothing is charged before day 8.
18+ · Bet responsibly. Only bet money you can afford to lose.
Put the sharp reference price to work.
Real-time Pinnacle odds, graded drops and your own Telegram alerts. 7 days free.
Start free trial