Football has traditionally been judged by a simple set of numbers: goals, points, wins, draws and losses. These figures remain fundamental, but they do not always tell the complete story of what happened during a match.
A team can lose 1–0 after creating several excellent scoring opportunities, while another team can win the same type of match with a single low-quality shot that happens to find the net. The final score records the result, but it does not necessarily describe the quality of the opportunities created by either side.
This is one of the reasons why Expected Goals, commonly abbreviated as xG, has become such an important statistic in modern football analysis.
Expected Goals provides a statistical way of estimating the likelihood that a particular shot will result in a goal. Instead of treating every shot as equal, an xG model attempts to distinguish between different types of scoring opportunities.
A shot from close range with a clear view of goal is obviously different from a speculative attempt from 30 metres. Traditional shot statistics count both as one shot. An xG model assigns them different probabilities.
Understanding this difference can provide a much deeper view of attacking performance, chance creation, finishing and defensive effectiveness.
What Does Expected Goals (xG) Mean?
Expected Goals is a statistical measure that estimates the probability of a shot becoming a goal.
An individual shot is generally assigned an xG value between 0 and 1.
A value close to 0 represents a shot with a very small probability of becoming a goal, while a value closer to 1 represents a very high-probability scoring opportunity.
For example:
- A shot with an xG of 0.05 has an estimated 5% probability of becoming a goal.
- A shot with an xG of 0.20 has an estimated 20% probability.
- A shot with an xG of 0.50 has an estimated 50% probability.
- A shot with an xG of 0.80 has an estimated 80% probability.
Important: This does not mean that a 0.80 xG opportunity must result in a goal. Football is a game of individual events and uncertainty. A player can miss a very good opportunity, while another player can score from a shot that historically has a very low probability of success.
The number represents a probability derived from comparable historical situations, rather than a prediction that a particular shot will definitely go in.
Why Was xG Created?
One of the fundamental problems with traditional football statistics is that goals are relatively rare events.
A team might have 15 shots and score twice. Another team might have five shots and score three times.
If we only look at goals, the second team appears to have been much more effective. But what if the first team created four excellent opportunities while the second team scored from three difficult attempts? The final score cannot answer that question.
xG attempts to add another layer of information. Rather than asking only "How many goals did the team score?", analysts can also ask: "How good were the opportunities that the team created?"
How Is xG Calculated?
An xG model is essentially a statistical model trained using historical football data. The model looks at large numbers of previous shots and examines which characteristics were associated with goals.
Different providers use different models, datasets and variables, which means that there is no single universal xG number. One provider might give a particular shot an xG of 0.31, while another model could rate exactly the same opportunity at 0.27 or 0.35.
Which Factors Influence xG?
1. Distance From Goal
Distance is one of the most obvious factors. A shot taken from six metres is generally more likely to result in a goal than an attempt from 30 metres. As the distance increases, the probability of scoring usually decreases.
2. Angle of the Shot
The angle from which the player shoots is also important. A player facing the centre of the goal generally has a better scoring opportunity than a player shooting from an extremely tight angle near the byline.
3. Body Part Used
Headers and shots taken with the feet can have substantially different scoring probabilities. A close-range shot with the foot may have a different historical conversion rate from a header taken from a similar position.
4. Type of Assist or Previous Action
The action that creates the shot can also provide important information. For example, a shot following a low cross into a dangerous central area may represent a different type of opportunity from a long-range shot following a loose ball.
5. Position on the Pitch
Where the shot occurs is central to xG modelling. A shot inside the six-yard box generally represents a substantially different opportunity from a shot outside the penalty area.
6. Penalties
Penalty kicks are a special situation because they are taken from a fixed location under relatively consistent conditions. Many xG models assign penalties a relatively high and relatively stable value (around 0.76–0.78 xG).
7. Defensive Pressure and Context
More advanced xG models can incorporate additional contextual information including goalkeeper position, defender positions, pressure on the shooter, shot height, and type of attack.
A Simple Example of xG
Imagine that Team A creates four shots:
| Shot | xG |
|---|---|
| Long-range attempt | 0.04 |
| Header from a difficult angle | 0.08 |
| Shot inside the penalty area | 0.25 |
| Close-range opportunity | 0.60 |
The team's total xG would be: 0.04 + 0.08 + 0.25 + 0.60 = 0.97 xG
This does not mean Team A is guaranteed to score approximately one goal in that particular match. It means that, based on the probabilities assigned to those opportunities, the combined expectation is approximately 0.97 goals.
What Does 2.00 xG Mean?
Suppose a team finishes a match with 2.00 xG. This does not mean the team "should" have scored exactly two goals. Instead, it means the total probability of the chances created corresponds to an expected value of approximately two goals.
If the team scores four goals, that is possible. If it scores zero, that is also possible. The difference between the expected value and the actual outcome is part of the natural variation in football.
xG For and xG Against
When analysing teams, two important concepts are:
- xG For: The expected goals generated by a team from its own chances. A higher value generally indicates that the team has been creating more or better scoring opportunities.
- xG Against: The expected goals generated by opponents against the team. A lower xG Against generally indicates that opponents have been restricted to fewer or lower-quality opportunities.
Together, these numbers provide more context than simply looking at goals scored and conceded.
Goals Minus xG
Another useful measurement is the difference between actual goals and expected goals: Goals − xG
A player with 15 goals and 11.5 xG has a difference of +3.5, meaning he has scored 3.5 more goals than the expected value of his chances. A player with 10 goals and 15.0 xG has a difference of −5.0, meaning he has scored five fewer goals than the expected value of his opportunities.
xG Per Shot
Another useful measure is xG per shot, calculated as: Total xG ÷ Total shots
A team with 10 shots and 2.0 xG has an xG per shot of 0.20 — the average shot carried an expected goal probability of 20%. A team with 20 shots and 1.2 xG has an xG per shot of 0.06 — its average opportunity was considerably weaker.
The Limitations of Expected Goals
xG is powerful, but it is not perfect:
- Different providers use different models — there is no universal xG formula.
- xG does not capture everything — it may not fully capture movement before the shot, defensive positioning, tactical decisions, or psychological factors.
- xG does not measure luck perfectly — random events are part of football.
- xG does not determine who deserved to win — matches are decided by actual goals, not expected goals.
- A high xG total is not automatically good process — analysts should investigate how the xG was generated.
Non-Penalty xG (npxG)
Non-penalty xG (npxG) removes penalty opportunities from the calculation. The purpose is to focus more directly on chances generated during open play and other non-penalty situations. This can make player comparisons more informative.
Why xG Matters
Expected Goals has changed the way many people analyse football. Its biggest contribution is simple: not all chances are equal.
Counting shots tells us how frequently a team attempted to score. Counting goals tells us what actually happened. xG adds another dimension by estimating the quality of the opportunities behind those attempts.
Used correctly, it can help football analysts understand attacking performance, chance creation, finishing, defensive quality, goalkeeper performance, tactical approaches, player evaluation, and differences between results and underlying chance quality.
The best way to use xG is not to ask: "What does xG say will happen?"
A better question is: "What does the quality and distribution of these chances tell us about how the team or player is performing?"
Frequently Asked Questions About xG
xG stands for Expected Goals. It is a statistical metric that estimates the probability that a particular shot will result in a goal.
A 0.25 xG shot represents an estimated 25% probability of becoming a goal according to the relevant xG model.
A total of 1.5 xG means that the combined expected value of the team's scoring opportunities was 1.5 goals. It does not mean the team was guaranteed to score one or two goals.
No. xG measures the expected value of scoring opportunities. It is not automatically a prediction of the exact final score.
Because xG represents probabilities rather than guaranteed outcomes. A team can create better chances but fail to convert them, while its opponent can score from fewer or lower-probability opportunities.
Yes. It can help evaluate the quality and quantity of scoring opportunities a player receives, although it should be combined with other statistics.
No. Different companies use different datasets and statistical models, so xG values can vary between providers.
Not directly. xG primarily evaluates the probability associated with chances. Comparing actual goals with xG can provide useful information about finishing outcomes, but it should not be treated as a complete measure of finishing skill.
Because it provides information about chance quality that traditional statistics such as goals and shot counts cannot provide on their own.
Sources and Further Reading
The explanations in this article were independently written and are not reproduced from the sources below. These references are provided so readers can explore the statistical and methodological background of Expected Goals.
- Opta Analyst — What Is Expected Goals (xG)?
- Opta Football Statistics Definitions
- Hudl StatsBomb — Expected Goals Explained
- PLOS ONE — Expected goals in football: Improving model performance
- BBC Sport — Expected goals: What is xG and how does it work?