Expected goals explained simply
What xG measures, why it tells the story of a match better than the result, and how to read it on a team page.
Read →How much a starter's absence weighs and how the news Radar updates the match numbers.
When a starter is missing, especially a striker or the goalkeeper, the quality of the chances a team creates or concedes changes. The model translates this into different expected goals and therefore different probabilities.
Take a match in which the home team has 1.6 expected goals and the visitors 1.1. With the Poisson model a home win is worth about 49% and Over 2.5 about 51%.
The news breaks: the home team's first-choice centre-forward won't play. If the model lowers its expected goals from 1.6 to 1.35, the home win drops to about 43% and Over 2.5 to about 44%. A single story, six or seven percentage points of difference.
A player back on the bench weighs little, the absence of the top scorer a lot. That's why the Radar shows how many points each probability changes and notifies you only above the threshold you choose.
Every story shows the source and the link to the original article. Official statements from clubs and disciplinary bodies are confirmed; rumours stay unverified until confirmation arrives.
Find out how the news Radar works.