What Is Expected Value (EV)?

What Is Expected Value?

Expected value measures the average profit or loss you would expect if you repeated the same type of bet under the same conditions many times.

Put simply, it tells you:

It is not about whether one individual bet wins or loses.

It asks whether repeating the same decision 100, 1,000, or 10,000 times would be likely to produce a profit over the long term.

The Basic EV Formula

Expected value can be calculated with the following formula.

To express the result as a percentage, multiply it by 100.

Suppose a bet is offered at odds of 2.00 and you estimate its true probability of winning at 55%.

The calculation is:

The EV of the bet is therefore:

This means that, over a large number of equivalent bets, the expected profit per bet is 10% of the amount staked.

It does not mean that every bet will return a 10% profit.

The first bet may lose.
You may even lose five in a row.

But if the probability estimate is accurate, the average return should converge towards the theoretical expected value as the number of bets increases.

Positive-EV Bets Can Still Lose

This point is essential.

A positive-EV bet is not a bet that is guaranteed to win.

Consider a bet at odds of 2.00 with a 55% probability of winning.

The bet has positive expected value.

It still loses 45% of the time.

In other words, it can lose on nearly one in every two occasions.

Even so, if the true win probability is higher than the probability implied by the bookmaker's price, the bet can be profitable over the long term.

Conversely, a bet that wins 80% of the time can still have negative expected value if the odds are too low.

For example, consider an 80% win probability at odds of 1.20.

The EV is:

Even though the bet is highly likely to win, that price may produce a loss over the long term.

In short:

A High Hit Rate Does Not Necessarily Mean Profit

Hit rate often receives a great deal of attention in sports betting.

Eight winners from ten bets.
An 80% hit rate.

On its own, that sounds impressive.

But if every bet was placed at odds of 1.10, one loss can wipe out the profit from several winners.

Meanwhile, a bettor with a 45% hit rate could still be profitable if the average odds were 2.50.

The hit rate alone is not what matters.

You also need to ask:

• At what odds was the bet placed?
• What was the true probability of winning?
• Did the market underestimate that probability?
• Did the bet have a structure that could remain profitable over time?

The hit rate is one part of the result. Expected value is a way to evaluate the quality of the decision.

Odds Are a Price, Not Just a Multiplier

Many people think of odds primarily as a payout multiplier.

At odds of 2.00, a ¥10,000 stake returns ¥20,000.
At odds of 1.50, a ¥10,000 stake returns ¥15,000.

That is not wrong.

But to understand expected value, you need to view odds as a price.

In investing, an excellent company can still be a poor investment if its share price is too high.

Sports betting works in much the same way.

You do not back a team simply because it is strong.

However strong a team may be, there is no value if the odds are too short.

Conversely, even a team with a low chance of winning may offer value if the odds are longer than its true win probability implies.

Bookmaker Probability vs Your Estimated Probability

Odds reflect the probability implied by the bookmaker's price or the wider market.

The basic formula for converting odds into implied probability is:

At odds of 2.00:

The implied market probability is 50%.

At odds of 1.50:

The implied probability is approximately 66.7%.

If the market implies a 50% probability but your analysis estimates a 55% chance of winning, that difference may represent value.

This probability gap is the source of expected value.

However, bookmaker odds build in a margin in the bookmaker's favour.

Simply converting the listed odds into probabilities therefore does not produce perfectly fair probabilities.

A more accurate analysis requires calculating fair odds after removing the margin.

We will explain fair odds in detail in the next article.

Why Do So Many People Bet on Negative EV?

There are several reasons why bettors are drawn to negative-EV bets.

1. Overvaluing Famous Teams

Well-known teams such as Real Madrid, Manchester City, Bayern, and PSG attract heavy public interest.

As more money backs a popular team, its odds can become shorter than they should be.

The team may still be likely to win, but that does not mean it is worth backing at the available price.

2. Focusing Only on Recent Results

When a team has won its last five matches, many bettors assume it will win again.

But unless you examine the quality of the opposition, home and away conditions, red cards, expected goals, and the role of chance, those results may not accurately represent the team's true ability.

3. Remembering Winners More Than Profitability

People tend to remember how often they won more readily than their overall return.

Winning several short-priced bets can create the feeling that the decisions were sound.

One large loss, however, can erase all the profit earned from those winners.

4. Failing to Compare Odds

The same bet can be priced differently by different bookmakers.

The difference between 1.80 and 1.90 may appear small.

Over the long term, it is significant.

If you care about expected value, the price you take matters just as much as the accuracy of your prediction.

What Do You Need to Find Positive EV?

The formula itself is not the hardest part of calculating expected value.

Doing that requires a wide range of information.

• Team attacking and defensive strength
• xG and xGA
• Total shots and shots on target
• Home and away differences
• Matchup history and tactical fit
• Unavailable players
• Fixture congestion
• Formations
• The importance of the match
• Market-specific characteristics
• Odds movement
• Sample size
• Data reliability

These factors must be evaluated together to estimate the win probability.

No amount of data, however, can predict the future perfectly.

Football contains many sources of uncertainty, including red cards, penalties, refereeing errors, injuries, weather, and finishing variance.

A win probability is therefore an estimate, not a certainty.

The objective is not to create a prediction that never loses.

AI Does Not Automatically Guarantee Expected Value

AI can help organise match data, analyse trends, calculate probabilities, and run simulations more efficiently.

But a bet does not have expected value simply because the underlying prediction came from AI.

Even if an AI model predicts a home win, the bet may not be worthwhile if the odds are too short.

The important question is:

AI is not a tipster.

Used properly, it can become an analytical engine that organises data, supports human judgment, and improves probability estimates.

EV Bet Engine is not being developed to turn AI into a match predictor.

It is being built to combine data, probability, odds, and risk so that we can judge whether a bet is worth making.

You can read more about the background and philosophy in Why we are building EV Bet Engine.

If EV Is Positive, Should You Always Bet?

Even when a bet has positive EV in theory, it does not follow that every such bet should be placed.

Examples include situations where:

• The estimated probability rests on weak evidence
• The data sample is small
• Injury or starting-lineup information is unavailable
• The market has low liquidity
• The odds have already fallen
• The model performs poorly in that market
• The estimate depends on several uncertain assumptions
• The bet is strongly correlated with other positions

Even when the calculated EV is high, the practical value may be limited if the underlying probability estimate is unreliable.

That is why EV Bet Engine considers more than expected value alone.

• Confidence
• Market characteristics
• Data quality
• Odds movement
• Risk
• Stake size
• Correlation with other bets

Positive EV is one condition for considering a bet.

It is not the only condition for approving one.

Do Not Confuse Expected Value with Short-Term Results

A good bet can lose.

A bad bet can win.

When a positive-EV bet loses, that does not necessarily mean the original decision was wrong.

When a negative-EV bet wins, that does not necessarily mean the decision was right.

You cannot evaluate the quality of an analysis from a single result.

The question to ask is:

The process should be reviewed, not just the outcome.

If losses continue, the model and its assumptions should of course be re-examined.

But if you keep changing direction based only on short-term wins and losses, you will struggle to maintain a strategy with genuine expected value.

Long-Term Success Requires Judgment, Not Just Prediction Skill

Long-term profitability in sports betting does not require the ability to predict the future perfectly.

It requires a repeatable process:

• Estimate probabilities
• Evaluate odds as prices
• Identify differences between your estimate and the market
• Manage risk
• Choose an appropriate stake
• Record results
• Continue reviewing each decision

Expected value sits at the centre of this process.

Instead of asking only which side will win, ask:

That is the first step toward treating sports betting as a problem of probability and decision-making rather than pure gambling.

Summary

Expected value, or EV, measures the average profit or loss you would expect if you repeated the same type of bet under the same conditions over the long term.

The basic formula is:

A positive EV indicates that the bet can theoretically produce a long-term profit.

Positive-EV bets can still lose.

What matters is not the outcome of one individual bet.

• Do not judge a bet by hit rate alone
• Treat odds as a price
• Compare market probability with your estimated probability
• Assess the reliability of the probability estimate
• Evaluate performance over the long term

Understanding expected value is the starting point for every form of serious sports-betting analysis.

In the next article, we will explain fair odds: the true price after removing the bookmaker's margin from the market odds.