Guide · 10 min read
EV Betting
Complete guide to expected value (2026)
By Lytic · Updated May 2026
Expected value estimates average return per unit staked from odds and win probability. Positive EV can still lose.
What is EV (expected value)?
Expected value is the mathematical average of an outcome over an infinite number of repetitions. In sports betting, EV describes the average result per €100 staked over many bets. It is not what happens on the next bet; it is what happens if you make the same bet 10,000 times.
The formula is simple, but the implications are profound. Let's walk through a concrete example. You are offered odds of 2.10 on a home win. The bookmaker's implied probability is 1 ÷ 2.10 = 47.6%. But you assess — using Pinnacle's lines (more below) — that the true probability of the home win is 52%.
That is a +EV bet with 9.2% edge. You do not know whether the home team wins this particular match — it might well lose. But if you consistently find bets with that type of mispricing, the expected result is positive. Actual results can still be negative for a long time, and the edge is only real if your probability estimate is right.
Why EV is the foundation of value betting
Most sports bettors lose money over time. One common reason is that they focus on outcomes, not on price. They ask "will it happen?" rather than "is the price right?"
Imagine a coin with a 52% chance of landing heads. Every time it lands heads you win €2. Every time it lands tails you lose €1. If you flip this coin 1,000 times, what might the journey look like? You could have a run of 50 flips where you lose 20 of them — it feels like you are playing it wrong. But you are playing it right. Random variance is mathematics' noisy short-term static; EV is the signal.
Judge the price, not the result of one bet. What matters is whether your average EV per bet is positive and whether your bankroll can absorb the swings. Even then, a positive expected value does not guarantee an actual profit.
It is also why discipline matters more than sporting knowledge. A bettor who understands EV and stakes systematically has a better process than one who "knows football" but stakes on gut feeling.
How to calculate true probability
The EV formula is simple. The hard part is knowing what the true probability actually is. This is where most ambitious bettors fail — they estimate probabilities from gut feeling or league tables and think that is enough.
A common shortcut is simpler and more reliable: Pinnacle no-vig. Pinnacle is a low-margin bookmaker widely used as a market reference, and its odds with the margin removed are a good estimate of the true probability.
What is implied probability?
Every set of odds can be converted directly to a probability. Odds 2.00 means the bookmaker prices the chance at 50%. Odds 3.00 = 33.3%. But bookmakers always add their margin (vig/overround), meaning the sum of all implied probabilities in a market exceeds 100%.
Odds 2.10 → 47.6% · Odds 3.50 → 28.6% · Odds 4.00 → 25.0% → Total 101.2%
To remove the vig and get the true (no-vig) probabilities, normalise: divide each implied probability by the sum of all of them.
Step 1 — Implied: 1/2.10 = 47.62% · 1/3.50 = 28.57% · 1/4.00 = 25.00% → Total 101.19%
Step 2 — No-vig: 47.62/101.19 = 47.06% · 28.57/101.19 = 28.24% · 25.00/101.19 = 24.70%
Step 3 — Compare: Your book offers odds 2.20 on home win → Implied 45.45%. Pinnacle no-vig: 47.06%. You are paying a price that underestimates the chance — that is +EV.
That is the core of the process. You do not need a statistical model of your own; the reference price does most of the work. Your job is to find bookmakers that deviate from Pinnacle in your favour — and bet quickly before their lines correct.
EV vs ROI — why EV is a better measure
ROI (return on investment) is the measure most bettors use to assess whether they are making money: (profit ÷ total stakes) × 100. It is a logical measure — but it is a lagging measure with one critical weakness. ROI tells you whether you have won money. It does not tell you whether you deserved to win it.
With 200 bets a skilled EV bettor can have negative ROI and a recreational bettor who got lucky can have positive ROI. Variance dominates in the short term. Statisticians say you need 1,000–2,000 bets before ROI gives statistically meaningful information about your edge.
| Metric | Reliable after | Measures |
|---|---|---|
| ROI | 1,000–2,000 bets | Actual result |
| Average EV | 100+ bets (early read) | Edge per bet |
| CLV | 100+ bets (early read) | Leading indicator |
Closing Line Value (CLV) is a useful leading indicator. CLV measures whether your placed odds were better or worse than the market closing odds at kick-off. If you regularly bet at better odds than the closing line, you have positive CLV. Research and bettor experience suggest positive average CLV against a sharp close is linked to positive long-run ROI, but commission, limits and voids can still turn it negative.
How to find positive EV in practice
The theory is clear. In practice EV betting is a workflow: fetch odds from your target bookmaker in real time, compare against Pinnacle's current line (adjusted for no-vig), calculate EV, and bet if EV exceeds your threshold — quickly, before the line moves.
That is what makes manual EV betting difficult. Pinnacle's lines move constantly as sharp money and information flow in. A +EV edge may exist for 2–10 minutes before bookmaker XYZ adjusts their line. You either need to sit glued to a screen with calculators open, or use a tool that automates the process.
Pinnacle comparison as method
The foundation of systematic EV hunting is the Pinnacle comparison. Every time a soft bookmaker deviates from Pinnacle's fair price in your favour by more than their individual vig margin, a +EV bet likely exists.
Hunting manually? Note Pinnacle's current prices for your target sport as the reference, then compare them line by line with the bookmakers you use. Only use bookmakers licensed where you live. If Pinnacle offers odds 1.90 on a home win (no-vig ~52.6%) and your book offers 2.05 (implied 48.8%), that is a clear +EV bet: (0.526 × 2.05) − 1 = +7.8%.
Lytic automates the entire process
Lytic scans hundreds of matches in real time, calculates EV using Pinnacle no-vig as reference (Shin de-vigging for 1X2 markets), ranks signals by edge size and shows Kelly stake based on your actual bankroll. CLV is logged automatically per bet. You see which bookmakers consistently deviate from Pinnacle in your favour — and can prioritise them before they sharpen their lines.
Get started with Lytic →Common mistakes EV bettors make
Understanding EV theory is one thing; applying it with discipline is another. These are the most common mistakes EV bettors make — and which turn a mathematical edge into an actual loss.
✕Too small a bankroll relative to stake size
EV never guarantees profit. A correctly estimated positive edge improves expected results over repeated bets, but model error and variance can still produce losses. With a bankroll of €500 and stakes of €50 (10%) drawdown risk is significant. Use smaller, consistent units and only stake money you can afford to lose.
✕Ignoring Kelly Criterion and flat staking
Flat staking is intuitive but mathematically suboptimal. Kelly Criterion calculates the optimal stake based on your edge and odds — too large a stake exponentially increases drawdown risk, too small unnecessarily slows growth. Quarter-Kelly (25% of the Kelly output) is a common rule of thumb.
✕Chasing losses with bigger stakes
A losing run of 20 bets with positive EV is statistically expected and nothing to panic about. Responding by increasing stake size is the fastest way to destroy your bankroll. EV betting requires emotional discipline: act on process, not recent outcomes.
✕Betting too late — when the line has already moved
Odds move fast. A +3% EV edge you identify at lunch may be +0.5% or gone entirely by the time you place the bet in the evening. Timing is critical — every minute without action is a minute during which Pinnacle and sharps may have adjusted the line.
✕Using the wrong reference price
Not all odds aggregators are equal. Using an average from five soft bookmakers as reference is meaningless — they often price identically and all contain the same inefficiency. Pinnacle closing odds are the standard reference precisely because they represent sharp money, not recreational opinions.
Frequently asked questions about EV betting
Is EV betting legal?
Where betting is legal, yes. EV betting is a mathematical way of identifying underpriced odds — not a violation of any rule. Betting at bookmakers licensed where you live and using analytical tools is allowed. The only practical issue is that bookmakers sometimes restrict winning accounts, but that is a business decision of the bookmaker, not a legal problem for you as a bettor.
How large an edge do you need to be profitable?
A correctly estimated 1–2% edge is positive EV, but small enough that estimation error, commission and limits can erase it. Large edges are rare and disappear quickly. Return is measured per unit staked, not on your bankroll, and actual results can stay negative for long stretches.
Is EV betting a source of income?
Very few people earn a living from it, and we do not present betting as income. Even with a real edge, results swing for months, accounts get limited, and the expected return on a small average edge is low. Treat betting money as money you can afford to lose.
Which sports have the best EV edges?
Football (Premier League, La Liga, Bundesliga, Champions League) dominates by volume and has the most liquid markets — meaning Pinnacle closing odds are highly reliable as reference. Tennis (ATP/WTA) is excellent for fast line movement. Ice hockey (NHL) and basketball (NBA) also work well. Avoid thin markets like lower divisions late at night — reference prices are unreliable and the risk of data errors is higher.
How long does it take to see positive ROI?
ROI needs 1,000–2,000 bets to settle — with 100 bets per month that means 10–20 months, and there is no guarantee it ends up positive. Closing Line Value (CLV) settles faster than ROI, but how fast depends on your average edge and odds. Treat 100 bets as an early read and several hundred as a firmer one. Track both.
What is the difference between EV betting and arbitrage?
Arbitrage means covering all outcomes at different bookmakers so the maths returns the same profit in theory — but operational risks apply (odds movement, stake limits, account restrictions). EV betting means staking on a single outcome you believe is underpriced, with full exposure to the result. Arbitrage margins are small (0.5–2% per arb) and only hold if every leg stands; EV betting can carry a larger expected edge per bet but comes with full variance. Arbitrage is also easier to identify automatically, while EV betting requires a reliable reference price such as Pinnacle.
Ready to start EV betting?
Lytic identifies positive EV in real time, calculates Kelly stakes automatically and tracks your CLV per bet — so you can measure your edge without waiting for 1,000 results.
Create your account →Related guides and tools