What Is Expected Goals (xG)?
The most important stat in modern football — and the one that's most misunderstood.
Expected goals (xG) is a statistical measure of the quality of a chance. It assigns a probability to every shot based on factors like distance, angle, and the type of assist. A shot with an xG of 0.3 would be scored roughly 30% of the time.
How xG is calculated
An xG model is trained on hundreds of thousands of historical shots. For any new shot, it asks: of all the shots that looked like this one, what share went in? The main inputs are:
- Distance from goal — the single strongest factor
- Angle — central shots are worth far more than shots from wide
- Type of assist — through-balls, cut-backs and crosses create very different chances
- Body part — headers are converted less often than shots with the foot
- Type of play — open play, set piece, counter-attack, penalty
Different providers use different models and inputs, so the same match can carry slightly different xG figures depending on where you read it.
What the values mean
| xG value | Meaning |
|---|---|
| 0.01–0.10 | Very low quality — a speculative effort, rarely scored |
| 0.10–0.25 | Low quality — scored occasionally |
| 0.25–0.50 | Medium quality — a real chance |
| 0.50–0.80 | High quality — more likely to be scored than not |
| 0.80+ | Very high quality — close to a certainty; a penalty sits around 0.76–0.79 |
xG vs actual goals
If a team's xG is consistently higher than its goals, it is underperforming its chances. If it scores more than its xG suggests, it is overperforming — which tends not to last.
Over one match, the gap is meaningless. Over ten or twenty, it becomes one of the better indicators of whether results are about to change direction.
Why xG matters
- Performance evaluation — it separates how a team played from how the scoreline finished
- Player evaluation — it distinguishes a striker who gets into good positions from one who scores spectacular one-offs
- Tactical analysis — it shows where chances are created and conceded, not just how many
- Recruitment — shot quality is far more repeatable season to season than conversion rate
The limitations
xG doesn't tell you everything. Most public models do not know where the defenders and the goalkeeper were, how much pressure the shooter was under, or what the game state was. It says nothing about the passes that never became shots, which is where a lot of good football lives.
Common questions
What is xG in football?
Expected goals — a measure of chance quality that gives each shot the probability of being scored based on where and how it was taken.
How is xG calculated?
A model trained on large numbers of historical shots estimates the scoring probability of a new shot from features like distance, angle, body part, assist type and phase of play.
What is a good xG for a match?
Around 1.5 expected goals is a typical team total. Anything above 2.5 usually means a side created several high-quality chances rather than many speculative ones.
What is the xG of a penalty?
Roughly 0.76 to 0.79 in most models, reflecting the long-run conversion rate of penalties in professional football.
Is xG reliable?
Over a run of matches, yes — it predicts future scoring better than past goals do. Over a single match it is a description of chance quality, not a verdict on who deserved to win.
See the numbers in a season
Gaffa gives you a full season of matches, chances and results to pick apart. Free, in your browser, nothing to install.
FM26 data and analytics → · Getting into football analytics → · What is a low block? →