StatGol

What xG means in football — and what it does not

xG stands for expected goals. It is the sum of the goal probability of every shot a team has created.

If a team has an xG of 1.8, it means the chances it created would, on average, be worth almost two goals. Not that it deserved two. Not that it is going to score two. That chances of that kind, repeated many times, would yield that many goals on average.

How the number is built

Each shot is given a probability of becoming a goal, estimated by a model from the characteristics of the attempt: the distance, the angle, whether it was a foot or a header, the type of move it came from. A penalty sits around 0.76 — three in four go in. A shot from outside the box at a tight angle sits at a few hundredths.

Add up the probability of all the team’s shots and you have the xG for the match.

What varies, and why that matters

The model belongs to whoever calculates it. There is no official xG. Different models are trained on different data, with different variables — some account for the position of the keeper and the defenders, some do not.

The practical consequence: the same match can have an xG of 1.4 from one source and 1.7 from another, and neither of them is wrong. Comparing xG across different sites is comparing two different rulers.

StatGol does not calculate xG: it arrives ready, carrying the model of whoever calculates it. That is stated here because blurring our own arithmetic with what is not ours would be the first way to mislead without lying — and the list of what the app computes and what it only displays is in the column reference.

What xG does not mean

It is not a forecast. It describes the quality of the chances that have already happened. No part of the number looks forward.

It is not a debt. “The team has 2.5 xG and no goal, the goal is coming” is the single most common misreading there is. xG has no memory: the next shot will carry whatever probability it carries, regardless of what did not go in before.

It is not merit. Creating the better chances and losing is a legitimate outcome of a football match, not a statistical injustice.

And it measures neither goalkeeping nor finishing. Two teams with the same xG may have had keepers on completely different nights. xG measures the chance, not the outcome, and not who wasted it — for what actually made the keeper work, the measure is shots on target.

Not every league has it

xG requires detailed tracking of every shot. Smaller leagues arrive without it — and when it arrives empty, StatGol shows a dash instead of a zero. Zero and “not measured” are different things, and showing a zero would be inventing a number.

See it live

On StatGol, xG sits next to shots, shots on target, dangerous attacks and corners, for every match in progress, refreshed every 30 seconds — which lets you see the gap between what a team has created and what it has scored while the match is still running.

Open the panel: 5 live matches are free, with no signup.

Read next: what “shots on target” means, what dangerous attacks are and the arithmetic behind every number.