What Is xG and How Do I Use It for Predictions?

xG (expected goals) is a statistic that measures the quality of a shot on a 0 to 1 scale, where the number represents the probability that an average player would score from that chance. Add up every shot in a match and you get a team's xG total - a strong indicator of how well a side actually played, regardless of the final score.

xG has become the most talked-about football statistic of the last decade. It appears on Match of the Day, in newspaper match reports, and on every major football data site. For predictors, understanding xG is one of the simplest ways to move from gut-feel guesses to evidence-based picks.

How xG Is Calculated

Every shot in a top-flight match gets an xG value based on a model that looks at the situation. We covered the full story in our piece on the xG revolution, but the short version is that the model considers things like:

• Distance from goal - shots from inside the six-yard box are worth far more than shots from 30 yards out
• Angle to goal - shots from tight angles are harder to convert
• Body part used - headers convert at a lower rate than feet
• Type of assist - through balls and crosses produce different chance qualities
• Defensive pressure - whether a defender is closing the shooter down

A penalty is roughly 0.76 xG. A tap-in might be 0.7 or higher. A speculative effort from outside the box is often 0.03 or lower. Add every shot up and you get the team total - a number that says, on average, how many goals this set of chances would produce.

Where to Find xG

xG numbers used to live in private analytics departments. These days they are everywhere. Match reports on the BBC and major newspapers usually include xG. FBref, Understat, and FotMob all publish full xG breakdowns for free. Most fantasy football and prediction tools surface xG alongside basic form data.

If you are predicting Premier League matches, you can typically find each team's xG-for and xG-against from the last six to ten games in under a minute. That is enough to inform most prediction decisions.

How to Read a High-vs-Low xG Game

Imagine a 1-0 win where the winning team had 0.6 xG and the losing team had 1.8 xG. The scoreline says one team won. The xG says the other team probably should have. That gap is the most useful thing xG tells you. Over a long enough run, results tend to drift towards the underlying numbers.