What is value betting? Expected value in football, explained

What value betting means

Value betting is not a system or a type of bet. It is a way of judging a price. Every price implies a chance - decimal odds of 2.50 assume a 40% chance, because 1 ÷ 2.50 = 0.40 (our guide to implied probability covers the conversion). A price is value when your estimate of the chance is higher than the chance the price implies. If you rate a home win at 50% and the price is 2.20, which implies 45.45%, the price is paying out as if the outcome were less likely than you believe it is. That is the whole definition.

Two things follow from it, and most of this article is about them:

• Value is relative to an estimate. It is only as good as the 50%. Change the estimate and the same price stops being value.
• Value says nothing about this match. A value price on an outcome that loses was still value. A bad price on an outcome that wins was still a bad price.

Thinking in value is the correct way to look at any price. It is not, by itself, a way to make money: for most people who try, the honest conclusion is that their estimates are not better than the market’s.

Expected value for a win-or-lose bet

Expected value (EV) is the average result per unit staked if you could make the same bet, at the same price, a very large number of times - assuming your estimate of the chance is right. For a bet that either fully wins or fully loses, with no refund or half settlement:

EV = p × odds − 1 (per unit staked)

Here p is your estimated chance as a decimal. If a bet can push, use EV = p(win) × (odds − 1) − p(loss); the refunded result contributes zero net profit. Quarter lines also need half-win and half-loss terms, as the quarter-line guide explains. The six examples below are all plain win-or-lose bets:

An EV of +0.10 means an expected net profit of 0.10 units for every unit staked over the long run. It does not mean 0.10 on this bet: on this bet you either win 1.20 units or lose the one you staked.

Short prices and long prices

The table shows why long prices are treacherous. At 8.00, every percentage point of chance moves the EV by 0.08 per unit; at 1.40, by 0.014. Rate the 8.00 shot at 14% instead of 11% - a three-point gap few people could defend - and the verdict swings from −0.12 to +0.12. A long-price estimate has to be far more precise to say anything at all, and it rarely is.

Short prices have the opposite trap. The outcome is likely, so it feels safe, but “likely” and “value” are different questions. At 1.40, any true win probability above 1 ÷ 1.40 (about 71.43%) gives positive EV. A true chance of 75% gives the 5% edge in the table; at the exact break-even probability the EV is zero. An estimate about 3.57 percentage points too high would erase that example’s edge.

Why a winning bet can be a bad bet

Football rewards bad decisions all the time. Take the 45% example at 2.20: an EV of −0.01. If the home side wins - 45 times in 100, on your own estimate - you collect 1.20 units and feel clever, but the decision was still slightly negative. The 50% example loses half the time, and when it does, nothing about the bet was wrong.

This is the core discipline of expected value betting: judge a decision by the price and the estimate at the moment you made it, never by the score. Over a single match the two can point in opposite directions, and they often do.

Variance: luck dominates small samples

Even a real edge takes a long time to show. Take the simplest case: every bet at 2.00, one unit staked each time, and a true 52.5% chance of winning - a 5% edge, since 0.525 × 2.00 − 1 = +0.05. Each bet wins one unit or loses one, so the swing per bet (the standard deviation of profit) is almost exactly one unit. Expected profit grows in step with the number of bets; the swing grows only with its square root.

The last column is the exact probability of finishing behind, assuming independent bets. After 100 bets, someone with a genuine 5% edge is still in loss more than one time in four, and the typical swing of ±10 units is twice the expected +5. The mirror image matters just as much: a bettor with a 5% disadvantage at the same price is in profit after 100 bets 27.4% of the time. At longer prices it is worse - at 8.00 the swing per bet is close to three units. A few dozen winning bets prove nothing about skill, and a few dozen losing ones prove nothing against it.

Why the market is hard to beat

The margin starts you behind

Every price carries the bookmaker’s margin. In the example 1X2 market from the implied-probability guide - 2.10 / 3.40 / 3.60 - the implied chances add up to 104.81%, so the average price is worse than fair. Remove the margin and the market’s own view of the home win is 45.4%, but the price of 2.10 needs 47.6% just to break even. If your estimates agreed exactly with the market’s fair chances, every one of the three prices would return 1 ÷ 1.0481 − 1 = −4.59% per unit. Agreeing with the market is not neutral; it is a steady loss. To find value, your estimate has to clear the margin’s share of the price - on that home win, about 2.2 percentage points above the market’s fair chance before you are even level.

The market aggregates information

A price, especially close to kick-off in a major league, combines many people’s models, opinions, money and team news, and the price that survives has been tested by everyone willing to bet against it. Your estimate - from a rating model, an expected-goals table, a feeling about motivation - is one more input, usually built on information the market already has. To be better, you need information the market has not priced yet or a better way of combining what everyone knows. Both are rare, and both fade as the market catches up.

Overconfidence

The most common source of apparent value is an estimate that is simply too confident. People remember the matches they read correctly, rate their own team with their heart, and round 48% up to “more than half”. A useful habit is to write your number down before you look at the odds, and keep the record. If your estimates and the market disagree often and by a lot, the likelier explanation is that your estimates are noisy, not that the market keeps missing things.

Closing line value: the benchmark

If results take a thousand bets to mean much, how can anyone tell whether their estimates are any good? The best-known answer is the closing line: the last price before kick-off. By then the market has absorbed the lineups and the late money, so the closing price is widely treated as its best pre-match estimate (our guide to dropping odds explains how a price travels there).

Closing line value (CLV) compares the price you took with that last price. Take 2.20 on a home win that closes at 2.00 and you beat the close by 10% (2.20 ÷ 2.00 − 1): by kick-off, the market thought your price had been too long. Consistently getting prices better than the close, over hundreds of bets, is the best-known evidence that an estimate has an edge. It is far less noisy than results, because it does not depend on what the ball did after kick-off.

The catch: the closing price still contains the margin

Beating the closing price by a little is not the same as beating the market’s closing estimate. In the 104.81% market above, a home price that closes at 2.10 has fair odds of 2.20 once the margin is removed. Taking 2.20 beat the closing price by 4.8% (2.20 ÷ 2.10 − 1) and still had an expected value of roughly zero against the market’s closing view. To be ahead of the close, your price has to beat the margin-free closing odds.

CLV is evidence, not proof: a few good prices are as noisy as anything else, and the closing price is itself an estimate, not the truth.

Common mistakes

• Confusing “likely” with “value”. A favourite at 1.25 will probably win; that says nothing about whether 1.25 is a good price for it.
• Favourite-longshot thinking. A big price is not a generous one. Long shots pay a lot because they rarely win, margins tend to be heavier on them, and small errors in the estimate swing their EV the most. The opposite belief - that favourites are “safe” - fails the same way: a short price leaves less room for error, not none.
• Trusting small samples. Thirty bets is almost nothing, as the variance table shows. A hot month is not evidence of an edge, and a cold one may not be bad luck.
• Chasing losses. Raising stakes to win back a loss changes nothing about the value of the next price; it only raises the cost of the next swing. A price is value or not regardless of what happened on the last bet.
• Calling it value after the move. By the time you notice a price has dropped, the drop is already in the new price.

Bankroll and responsible gambling

If you bet at all, decide in advance how much you are prepared to lose over a week or a month, and treat it as the cost of entertainment, not an investment. Never stake money you cannot afford to lose, never borrow to bet, and never raise stakes to recover losses. Set deposit and time limits, and set them on a calm day rather than a bad one. If betting stops being fun, or you find yourself hiding it, stop - free, confidential help is linked at the foot of this page.

The variance table is the honest reason for all of this: even someone with a real edge can lose for a long time, and most people who bet do not have one.

Where Match-Lens fits

Match-Lens takes no bets, gives no betting advice and does not tell you what is value. It shows how the market’s view moved. Every fixture keeps three moments of its 1X2 price:

• S - the first pre-match price we recorded;
• K - the last recorded price before kick-off, our closing snapshot;
• L - the live price, once the match is on.

Where K came in from S, the field behind it warms from orange towards red - by the change in chance, not the change in the number. On the Schedule, the Max drop column sorts fixtures by the biggest shortening between opening and last pre-kick-off price. Opening a match shows its odds window, with FULL TIME, HALF TIME and CORNERS pages: the S and K rows are the bookends, followed by the live tick history with the minute and score.

Results keep finished matches with their full-time statistics and their opening and kick-off prices, going back years, so you can compare a price you saw with where the market closed - across many matches rather than one. It is information, not advice.

Frequently asked questions

Read next

• Implied probability: how to read football odds as chances
• Asian handicap explained: every line from 0 to −2 with examples
• Dropping odds in football: what an odds drop means and how to read it