What is xG (Expected Goals) in Football?

If you've watched football punditry over the last decade, you've heard the phrase "expected goals" — usually shortened to xG. It's become one of the most widely used stats in the sport, but it's also one of the most widely misunderstood.

The basic idea

xG measures the quality of a chance, not whether it was scored. Every shot is assigned a number between 0 and 1, representing the probability that an average player would score from that exact position, under those exact circumstances.

A tap-in from a yard out might carry an xG of 0.9 — nine times out of ten, a professional finishes that chance. A speculative strike from 35 yards might be worth 0.02 — it goes in roughly one time in fifty. Add up the xG of every shot in a match and you get a match's total xG: a measure of the quality of chances created, independent of finishing.

How it's calculated

xG models are built by analysing tens of thousands of historical shots and the outcome of each one. The key inputs are usually:

• Distance to goal — closer shots score more often.
• Angle to goal — a shot from a tight angle is harder to convert than one from a central position, even at the same distance.
• Body part — headers convert at a noticeably lower rate than shots from the foot.
• Assist type — a through ball or a cutback tends to produce a better chance than a shot from open play with no clear buildup.
• Defensive pressure — how many defenders and the goalkeeper are positioned between the shooter and the goal.

Feed enough historical shots with these attributes into a statistical model, and it learns the scoring probability for a new shot with a similar profile.

Why it's useful

The scoreline tells you what happened. xG tells you what was likely to happen, which makes it a much better tool for answering questions like:

• Did a team deserve to win, or did they ride a couple of fortunate finishes?
• Is a striker's goal tally sustainable, or are they overperforming their underlying chances?
• Is a team creating good chances but wasting them, or creating very little and being clinical?

A team that wins 1-0 while being outshot 15-3 in high-quality areas has probably been flattered by the result — and xG is the stat that surfaces that, where the scoreline alone hides it.

What xG isn't

xG is a measure of chance quality, not a prediction of the final score, and it isn't adjusted for who's taking the shot — a 0.3 xG chance falling to Erling Haaland and the same chance falling to a weak-footed defender are treated identically by most public models. It's also just one metric: a team can generate excellent xG numbers while being tactically toothless in ways the stat doesn't capture, like failing to control midfield or defending set pieces poorly.

Used well, though, xG is one of the fastest ways to separate genuine performance from noise — and it's the foundation most of the stats on this site are built from.