What is xG? A practical guide to expected goals

What expected goals measure

xG estimates the probability of a shot becoming a goal, on a scale from zero to one, using past shots with similar characteristics. It describes chance quality rather than guaranteeing how many goals a team will score.

Why chances receive different values

Common model inputs include shot distance, angle, body part and how the chance was created. Providers can return different values because their models use different details. Identify the provider before comparing numbers.

Read the three-shot example

The table below is entirely hypothetical: it is not a real fixture or data from our app. Adding 0.10, 0.25 and 0.40 gives 0.75. This is not three quarters of an actual goal. Actual goals are whole numbers; those attempts can produce no goals, one goal or more.

Match xG is different from the score

Actual goals can exceed or fall below an expected-goals total. One match is insufficient to judge a team or model. Separate what happened on the scoreboard from the description of the chances, then look at several matches and their context.

What one number cannot tell you

A chances total alone does not explain why a coach changed formation or confirm player availability for the next fixture. Ask when the chances occurred, how the game changed and whether the numbers came from the same provider. A metric does not replace reading the match.

Use the concept when reading Forecast

Distinguish chance quality from a team’s probability of winning. Read the report summary, factors and timing rather than reducing the analysis to one number. This guide explains a football concept; it does not promise xG coverage for every competition or Forecast report, or imply a Hudl Statsbomb data partnership. You can request Forecast on the website and read issued reports in your account Vault.

Forecast questions