Expected Goals For
Expected goals for is the total quality of the chances a team has created, added up across every match of the season.
Why it matters. A higher figure means a better attack whether or not the finishing has arrived yet, which is why it moves ahead of the goals column.
Every shot, added up
Expected goals for is a running total: take the xG value of every shot a team has taken this season and sum it, match by match. A team creating 1.8 xG a game sits at roughly 68 across a 38-match season. No weighting, no adjustment for opponent, just the raw quality of every chance the attack produced.
Penalties count at full value. A penalty is a shot, it carries a base rate near 0.78, and the team's play won the foul that led to it. Stripping penalties out is the right move for judging an individual finisher, because it separates their open-play craft from a spot kick anyone could take. It is the wrong move for a team total, where the question is simply how much scoring opportunity the team's overall play generated. Own goals stay out on the other side of the ledger entirely: nobody on the scoring team took a shot, so there is nothing for the model to value.
Reading the gap to goals scored
A team's goals trail their xGF. A side sitting on 1.3 goals a match against an xGF of 1.9 a match is finishing below its chance quality. That is a squad creating the positions a good attack creates, with the conversion still to arrive. Watch the goals column catch up over the following weeks more often than the xGF column falls.
A promoted side starts hot on modest chances. A newly promoted team can post five wins from a start with an xGF nowhere near the division's better attacks. Real goals look excellent. xGF says the underlying chance creation has not yet matched it, and that gap is the one to track rather than the results.
The trend moves before the table does. Track xGF over a team's last six to eight matches rather than the season figure alone. An attack finding better positions shows up in a rising per-match xGF well before it shows up in three straight wins. That lead is the entire reason to look at xGF at all.
xGF settles whether the attack is actually good
Strip out the run of fixtures, the finishing streak and the scoreline noise. What is left is one question: is this team getting into good scoring positions, repeatedly, against the opposition it has faced. A side can win ugly for a month and still carry a strong xGF underneath it. A side can lose narrowly for a month and carry the same. The figure does not care which result actually happened. It cares what the chances were worth.
Where this gets misread
- "xGF is just goals scored, smoothed out." No. It is an independent measure of chance quality built from shot locations and types, not from whether the ball crossed the line. The two can and do diverge for a full season, not just a handful of matches.
- "A low xGF with a high goals total means the model is wrong." It usually means hot finishing. That is a real thing that happens to real teams, but a finishing rate well above what the chances support tends to fade rather than hold.
- "Penalties shouldn't count towards a team's attacking quality." They are shots, taken because the team's play won a foul in the box, and they carry a value like any other shot. Leaving them in is deliberate, not an oversight.
Most expected goals created this season
Common questions
- Does expected goals for include penalties?
- Yes. Penalties are shots and carry their expected value like any other. Own goals are excluded, because an own goal is not a shot and so has no expected goals attached to it.
- How does it compare against goals scored?
- Over a season the two usually land close together. A persistent gap points either to unusually good finishing or to a team creating a kind of chance the model values differently from the eye.
- Why do individual player stats strip penalties out if the team total keeps them in?
- Different questions. A team's xGF asks how many goals its overall play deserved, and a penalty won by that play counts. A player's non-penalty xG asks how good their open-play chances were, and a penalty says nothing about that, so it is removed.
- Can a team have a strong attack with a modest xGF?
- Yes, if it plays fewer matches at a high tempo or sits deep and counters. xGF is a season total, so a team creating good chances at a slower rate can still trail a side that creates more, lower-quality efforts in bulk. Check it alongside xGF per match, not instead of it.
- Does a low xGF always mean a weak attack?
- Usually, but check the run of matches first. A handful of games against well-organised defences can suppress it temporarily. A low figure held across ten or more matches is a structural read, not a blip.
- How quickly does xGF respond to a change in a team's play?
- Faster than goals do. A team creating better chances shows it in the shot data within a handful of matches, well before the goals column catches up. That lag is why xGF trend is worth checking before a run of results confirms it.
Related terms
See xGF across the Premier League
Every player, team and match measured on the chances created rather than the scoreline.
Last updated