What is xG, and what does it really tell you about a match?
Aug 17, 2026 · 13:25
xG is not a magic number: it measures chance quality - here is what you can and cannot read from it.
Expected goals, or xG, has become a standard statistic in football coverage. The idea is simple: every shot is assigned a value between 0 and 1 representing how often similar attempts have been converted in the past.
What determines the value of a chance?
Models typically start from the distance and angle of the shot, but they also account for the body part used, the type of chance (a counter-attack, a header from a cross, a penalty), the number and position of defenders and the goalkeeper's position. A penalty is worth roughly 0.75-0.79, while a shot from 20 metres through a crowd of players is worth a fraction of that.
What it is good for, and what it is not
xG is useful for showing whether a team consistently creates good chances or simply takes lots of poor shots. Over ten to fifteen matches it predicts future results more reliably than goals scored. It cannot decide whether a single result was deserved: a goal from a 0.3 xG chance is still worth three points.
The most common misunderstanding
If a team records 2.4 xG and fails to score, that is not necessarily bad luck. Perhaps the chances fall to poor finishers, or the model cannot see a defender's outstretched leg. Models also differ, which is why two websites can publish different xG values for the same match.
How to use it
Look at the gap: if a team keeps scoring more than its xG suggests, that usually corrects itself over time. xG is not a prediction but a filter that helps separate luck and form.
