xG vs xGoT: how to measure finishing
Two strikers get identical chances all season. One keeps passing the ball into the corner; the other keeps hitting the keeper. Classic xG can't see the difference between them, because it stops measuring the moment the ball is struck. That's the gap xGoT exists to fill, and together the two metrics let you take finishing apart into its real components.
Two measurements, two moments
xG (expected goals) values a chance before the shot: distance, angle, assist type, situation. It asks: how often does an average player score from here? What the shooter then does with the ball changes nothing. (The full xG explainer is here.)
xGoT (expected goals on target) values the shot after it's struck, using where the ball is headed in the goal frame: the same effort aimed at the top corner carries far more xGoT than one straight down the middle. Only on-target shots carry xGoT at all; a miss tells you nothing about placement because there's no placement to judge.
One chance, two numbers. A 0.3 xG chance arrowed toward the corner might become 0.8 xGoT; the same chance dragged softly at the keeper might become 0.05. The chance was identical; the strike wasn't.
The gap is the skill
Subtract one from the other, over a proper sample, and the components of finishing separate cleanly:
- xGoT minus xG is strike quality: how much a player's ball-striking adds to (or subtracts from) the chances he takes. Persistently positive means he makes chances harder to save than they arrived; that's placement and power, the purest shooting skill the data can see. FDscout carries this as the Strike quality +/- composite on player profiles.
- Goals minus xGoT captures what happens after the strike, which over big samples mostly reflects the goalkeepers faced (and is precisely how keeper shot-stopping gets measured, as xGoT faced minus goals conceded).
- Goals minus xG, the familiar headline gap that drives our best finishers ranking, is the sum of both parts: strike quality plus outcome.
This decomposition is where it gets practical. A striker beating xG through xGoT is genuinely striking the ball beautifully; expect it to continue. One beating xG without an xGoT edge has been finding keepers in bad form, and that regresses. Same headline overperformance, different futures.
What each metric misses
xG knows nothing about execution; xGoT knows nothing about shot selection, because a terrible decision struck sweetly still earns high xGoT. Read alone, each flatters a different flaw: xG flatters the wasteful shooter with good movement, and xGoT flatters the ambitious blaster who occasionally middles one. Read together with xG per shot (does he shoot from good positions in the first place?), the full shooting profile emerges: where he shoots from, how well he strikes it, and what it all turns into.
The usual honesty caveat applies double here: xGoT gaps are noisy, only on-target shots count toward them, and one season is a hint rather than a verdict. Volume and repetition are what turn these gaps into evidence.
Reading it on FDscout
Every player profile carries xG, xGoT, xG per shot and the strike-quality composite, each as a percentile against the same position across 30 leagues (how percentiles work). The related trait badges do the summarising for you: xG out performer for sustained scoring above expectation, Shoots from good positions for elite chance selection. Open any forward on FDscout and you can tell, in one look, whether the goals come from movement, from striking, or from both, and that's the difference between knowing a player scores and knowing how.