What expected goals (xG) is

Goals are rare and noisy, so the final score is a poor read on who actually played well. Expected goals (xG) was built to fix that, and it is the sharpest single number in football analytics.

The core idea

Every shot is scored by how likely it was to be a goal, from 0 to 1. A tap-in from two yards might be worth 0.8 xG; a speculative shot from 30 yards might be 0.03. Add up a team's shot values across a match and you get its xG — the number of goals an average finisher would have scored from those exact chances.

Why it beats raw goals

A side can win 1-0 on a deflected shot while creating almost nothing, or lose 1-0 having racked up 2.5 xG of clear chances. The scoreline says one thing; xG says the opposite — and over many matches, xG is the better guide to true quality. It strips out the luck of a single bounce.

How to read an xG line

A match might read home 2.1 xG, away 0.6 xG, final score 1-1. That tells you the home side dominated the chances and was unlucky not to win, while the away goal was against the run of play.

xG measures the quality of chances, not the result. A team that consistently out-xGs opponents is creating the conditions to win, even when the scoreboard disagrees.

For predictions, xG is the honest version of "who is good". It feeds directly into estimating how many goals a team should score next time — the basis of the goal models in the dedicated model track.