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 →Why a team's 70% Over 2.5 isn't enough on its own and which numbers to look at alongside it.
A team's Over 2.5 percentage tells you in how many of its matches at least 3 goals were scored in total. It's a useful number, but read on its own it often leads you astray.
Many teams play more open matches at home and tighter ones away. For a match what counts is the home team's Over in home games and the visitors' Over away, not the season total. It's the same principle as home advantage.
After 10 matchdays a team has played about 5 home matches. An 80% home Over can mean 4 matches out of 5: one different game is enough to drop it to 60%. The smaller the sample, the more cautiously the percentage should be taken.
A match has two teams. A team with 75% Over against one with 30% doesn't automatically produce an Over match: the two numbers need combining, and that's exactly what the model does when it estimates both teams' expected goals.
A team may have seen lots of goals thanks to one-offs: penalties, own goals, goalkeeping errors. Expected goals tell you whether there's a style of play behind those goals that produces them consistently.
You'll find every percentage, from 0.5 to 4.5, in each league's Over/Under statistics.