How to Calculate Expected Value (EV) in Football Betting

Expected value is the foundation of professional betting. It's the average profit or loss you can expect per bet if you repeat the same scenario many times.

Without understanding EV, you're betting blind. With it, you can quantify your edge, compare different bets, and measure whether you're actually profitable.

The Expected Value Formula

The simplest formula for expected value in betting is:

EV = (Probability of Winning × Odds) - 1

Or expressed differently:

EV = (Probability of Winning × Profit) - (Probability of Losing × Stake)

Both formulas are equivalent. The first is faster for single bets. The second is more intuitive if you think in terms of profit and loss.

Formula Breakdown

If you bet £100 at 2.50 odds with a 45% probability of winning:

Using the first formula: EV = (0.45 × 2.50) - 1 = 1.125 - 1 = +0.125

This means for every £1 bet, you expect to profit £0.125 (12.5 pence) over many repetitions.

On a £100 bet, that's £12.50 expected profit.

Using the second formula:

If you win at 2.50 odds on a £100 stake, your profit is £150 (2.50 - 1 = 1.50 × £100). If you lose, you lose your £100 stake.

EV = (0.45 × £150) - (0.55 × £100) = £67.50 - £55 = +£12.50

That's a £12.50 expected profit per bet, matching the first formula (12.5% of the £100 stake).

The expected value is positive. This is a good bet.

Interpreting EV Results

Positive EV (+EV)

If EV is positive, the bet has an edge in your favour. Over many repetitions, you'll profit.

EV of +0.05 means 5% expected profit per bet. EV of +0.02 means 2%. Both are profitable. The larger the EV, the better the bet.

Negative EV (-EV)

If EV is negative, the bet is a losing proposition long-term.

A casual bettor might still win because variance allows short-term luck. But mathematically, negative EV bets lose money. Avoid them, regardless of confidence in the outcome.

Zero EV

A bet with zero EV is breakeven. You expect neither profit nor loss long-term. These are rare in betting because bookmakers charge a margin (the overround). Almost all bets offered to you are negative EV, which is why they make money.

Your job is finding the rare +EV bets.

Worked Example 1: Single Bet

Manchester City vs Newcastle. You assess City's probability at 72%. Bookmaker odds are 1.55.

Step 1: Convert odds to implied probability.

Implied probability = 1 / 1.55 = 0.645 (64.5%)

Step 2: Compare your probability (72%) to implied probability (64.5%).

Your assessment is higher. There's potential value.

Step 3: Calculate EV.

EV = (0.72 × 1.55) - 1 = 1.116 - 1 = +0.116

Step 4: Interpret.

+0.116 means 11.6% expected ROI. On a £100 bet, you expect to profit £11.60 long-term. This is a solid bet with meaningful edge.

Worked Example 2: Comparing Two Bets

You've identified two potential bets:

Bet A: Team to win at 2.20 odds. You assess 50% probability.

EV = (0.50 × 2.20) - 1 = 1.10 - 1 = +0.10 (10% ROI)

Bet B: Over 2.5 goals at 1.85 odds. You assess 60% probability.

EV = (0.60 × 1.85) - 1 = 1.11 - 1 = +0.11 (11% ROI)

Both are positive, but Bet B has slightly higher EV. If you can only bet one, choose Bet B.

If you can bet both, assign more capital to B because it has a larger edge, assuming your probability estimates are accurate.

Worked Example 3: Obviously Bad Odds

Man United to beat a lower-league team in the cup. You assess 95% probability. Odds are 1.05.

EV = (0.95 × 1.05) - 1 = 0.9975 - 1 = -0.0025

Negative EV. This is a bad bet, even though Man United is almost certain to win. The odds don't compensate for the actual probability. Avoid it.

EV for Accumulators

An accumulator (parlay) links multiple bets. You win only if all legs win.

Calculating EV for accumulators is more complex because the legs interact.

For two independent bets with odds O1 and O2, and probabilities P1 and P2:

Accumulator EV = (P1 × P2 × O1 × O2) - 1

Accumulator Example

Bet A: Team 1 to win at 1.80 odds. You assess 55% probability.

Bet B: Team 2 to win at 1.90 odds. You assess 55% probability.

Accumulator EV = (0.55 × 0.55 × 1.80 × 1.90) - 1

= (0.3025 × 3.42) - 1

= 1.034 - 1

= +0.034 (3.4% ROI)

The accumulator has positive EV. But notice the EV dropped from the individual bets (which are around 10% each) to 3.4% combined.

This is because accumulators compound probability. Winning both legs is less likely than winning either leg alone. The bookmaker's margin compounds too.

For this reason, accumulators are often -EV or barely +EV, even when individual legs are +EV. Professionals avoid them unless the EV is substantial.

Why Variance Means You Lose Sometimes

A +EV bet will lose. Sometimes. Even frequently.

If you have a +0.10 EV bet, you win roughly 55% of the time and lose 45%. That means you expect to lose 45 out of every 100 bets. Losing streaks happen. Variance is normal.

Example:

You place 10 bets, each with +0.10 EV and 55% win rate.

Expected outcome over 10 bets: 5.5 wins, 4.5 losses. Profit: roughly £55.

But you might get unlucky:

Run 1: 3 wins, 7 losses. Loss: £40.

Run 2: 8 wins, 2 losses. Profit: £80.

Both runs are possible. The long-term average is +£55. But short-term variance is significant.

This is why professionals place many bets and track over months or years, not weeks.

EV and Closing Line Value

EV is the absolute measure of edge in a single bet. But how do you know your probability estimate is actually correct?

Closing line value (CLV) is the test.

If you consistently beat the closing odds (the final price before kick-off), it indicates your probability estimates are accurate and you genuinely have +EV.

If you consistently lose to the closing odds, your estimates are inaccurate and your "edge" is illusory.

Track both EV and CLV. They're complementary.

Practical Application: Building a Betting Record

To apply EV in practice:

1. For each bet, record:

   • The date and match.
   • Your probability estimate.
   • The odds you backed.
   • Implied probability (1 / odds).
   • EV calculation.
   • The outcome (win or loss).
   • The closing odds (the price at kick-off).
   • Closing line value (whether you beat them).

2. Every 100 bets, review your record.

3. Calculate your actual win percentage.

4. Compare to your predicted win percentage (based on probability estimates).

5. If actual is significantly lower, you're overestimating probability. Adjust downward.

6. If CLV is consistently negative, your timing or value identification is poor.

This feedback loop improves your estimates and validates whether you genuinely have an edge.

Why Most Bettors Don't Calculate EV

EV requires you to estimate probability, and most bettors overestimate their ability to do so. They don't want to calculate it because they'd realise many of their bets are -EV.

It's easier to back your favourite team, win, and feel smart. Harder to acknowledge that even when you win, you might have been lucky, not skilled.

Professionals force themselves to calculate EV. It's uncomfortable. But it's the only path to profit.

In Summary

• Expected value (EV) is the average profit or loss per bet over many repetitions, calculated as: EV = (Probability × Odds) - 1
• Positive EV indicates an edge in your favour; negative EV is a losing proposition long-term, regardless of short-term outcomes
• A bet with +0.10 EV means 10% expected profit per bet; bets are good or bad based on EV, not on whether they win or lose
• To find +EV bets, estimate the true probability of an outcome and compare it to the implied probability from the bookmaker's odds
• EV and CLV are complementary: EV is your theoretical calculation, CLV reveals whether your estimates are accurate by comparing to closing odds
• Accumulators compound probability and the bookmaker's margin, resulting in lower EV than individual legs, despite each leg being +EV
• Variance means +EV bets will lose; a +0.10 EV bet wins roughly 55% of the time, meaning 45 out of 100 bets lose on average
• Professional bettors track EV for every bet and review every 100 bets to validate whether estimated probabilities match actual win rates
• Most bettors avoid calculating EV because it reveals many of their bets are -EV; measurement beats guessing over time

Frequently Asked Questions

What's the minimum EV to be worth betting?

This depends on your bankroll and comfort with variance. Some professionals bet anything above +0.01 EV (1% ROI). Others require +0.03 EV (3%) to justify the risk. Start with bets where you're confident and EV is obvious, say +0.05 minimum. As you build a track record, you can bet tighter margins.

How accurate do my probability estimates need to be?

Very. An error of 2-3% in probability can swing EV from positive to negative. This is why tracking your record is critical. You'll discover your actual accuracy and adjust. Many bettors overestimate by 5-10%, turning apparent +EV into actual -EV.

Can I use EV for bets with uncertain probabilities?

Yes, but your EV calculation is only as good as your probability estimate. If you're guessing, your EV is unreliable. Only use EV for outcomes where you've invested time in analysis or modelling. Otherwise, you're quantifying a guess, which is worse than just guessing.

What about live betting and in-play markets?

EV applies to any bet. Live odds change constantly. Calculate EV for live bets the same way. Live markets can be less efficient, creating more value opportunities, but they also move faster so you need quick decision-making.

Should I always bet the highest EV bets?

In theory, yes, if you have unlimited bankroll. Practically, factor in variance. A bet with +0.30 EV might require large stakes, creating short-term volatility. A bet with +0.05 EV might have lower variance. Some professionals size bets by EV and bankroll using the Kelly Criterion to balance growth and stability.

What if my EV is +0.01 but I'm only 52% confident?

If you estimate a 52% probability and EV is +0.01, your probability estimate might be wrong. A 52% win rate has massive variance. You could lose 20 in a row on a true 52% proposition. Only back bets where you have reasonable confidence in the probability, say 55%+, unless you're placing extremely large volumes.