What is xG, and what can it not tell you?
How xG works
A shot from close range, straight in front of goal, scores far more often than a shot from 30 metres. xG puts a number on that. A chance rated 0.10 is the kind of chance that becomes a goal about once in ten attempts. A penalty is rated much higher.
A team’s xG for a match is the sum of its shots. If a side ends a game with 1.8 xG and the other with 0.6, the first created clearly better chances, whatever the score says.
Same xG, different story
Two teams can have the same total xG and very different chances of actually scoring:
Ten chances of 0.1 give at least one goal in about 65% of cases (1 − 0.910). Two chances of 0.5 give it in 75% (1 − 0.52). The total hides the shape.
What xG can’t tell you
- Finishing and goalkeeping. xG describes the chance, not the player taking it or the keeper facing it.
- Game state. A team leading 2–0 often sits deep and lets the opponent shoot. Its low xG may be a choice, not a weakness.
- Chances without a shot. A dangerous cross that nobody reaches has no xG at all.
- Small samples. Over one match, xG and the score often disagree. Over a season the picture is steadier.
- Every league. Not every competition has xG data. Where it’s missing, it simply isn’t there.
How Edgity uses xG
When both teams have xG from at least a few matches, the model blends it with actual goals to estimate how strong each attack and defence is. The more matches with xG, the more weight it gets, but it never fully replaces real goals.
The match page also shows the model’s expected goals for each team. When a match has no recorded xG, the page says so: the numbers you see are then the model’s estimate, not measured xG. In the web app, the live Pressure chart uses xG alongside shots and corners, where the league provides it.
The takeaway
xG is a good measure of chance quality and a poor predictor of a single score. Use it to understand how a team plays, not to decide what happened in one match.