What Is xG (Expected Goals) in Football? Meaning Explained

xG (expected goals) is a statistic that rates the quality of every shot on goal and estimates how many goals a team should have scored given the chances it created.

If a match finishes 1-0 but the stats show xG 0.4 against 2.1, it means the winning side barely created any danger, while the losing side dominated the quality of chances without converting them. xG separates the scoreline from how deserved it was, which is exactly why it has become a standard metric in modern football analysis.

What xG means in football

xG stands for expected goals. In football, the meaning of xG is the number of goals a team's chances were worth: every shot gets a value from 0 to 1, and the values add up over a match, a run of games or a season. An xG of 1.8 does not predict that a team will score 1.8 goals; it says that chances of this quality would produce about 1.8 goals on average if they were played out many times.

That makes xG a measure of how well a team creates and prevents chances, not of how well it finishes them. A striker who scores 15 goals from 9 xG has finished well above the average for those chances; a team that concedes 30 goals from 45 xGA has its goalkeeper or a run of luck to thank.

How xG is calculated

Every shot on goal is assigned a number between 0 and 1 — the probability that a goal is scored from that specific situation. The model is trained on tens of thousands of comparable shots from past matches and factors in:

• distance to goal and the angle available to the shooter;
• whether the shot is taken with the foot or the head;
• the type of pass that created the chance (cross, cutback, through ball, one-on-one);
• the position of defenders and the goalkeeper at the moment of the shot;
• whether the shot is a penalty — penalties get a fixed xG equal to the average conversion rate.

A team's match xG is simply the sum of the xG of every shot it took. A team that manages five shots from good positions inside the box can post a higher xG than an opponent who fires fifteen shots from distance under no pressure.

Under the hood, an xG model is a statistical model, such as logistic regression or gradient boosting, trained on past shots: it learns how often shots from similar situations were scored. Every data provider trains its own model on its own data, so the xG of the same match can differ slightly between sites. Compare numbers from one source rather than mixing them.

xG explained with an example

Take one team's shots in a match and rate them with the location-only model of the xG calculator described below:

• a penalty — 0.79;
• a central shot from the edge of the six-yard box — 0.28;
• three central shots from the edge of the penalty area — 0.06 each.

The team's xG for the match is 0.79 + 0.28 + 3 × 0.06 = 1.25. Without the penalty the same match is worth only 0.46 xG from four shots, most of them from distance — which is why non-penalty xG is often quoted separately. A data provider's model would give these shots slightly different values, because it also knows the body part, the pass before the shot and where the defenders were.

xG vs. shot count

Shot count says nothing about shot quality: a 25-metre effort under pressure and a tap-in from a cutback both count as one "shot" in the basic stats, but they look completely different in xG. That's why a team that shoots a lot from range often loses the xG battle to a team that shoots less often but from better positions. For more on reading shots, shots on target and corners together, see "How to Read Shots on Target and Corner Stats".

xG, xGA, npxG, xGOT and xPts: the related metrics

Once you know what xG in football means, the neighbouring abbreviations are easy to read:

• xGA (expected goals against) — the xG of the shots a team allowed. It describes defence the way xG describes attack: a side that concedes few goals but has a high xGA has probably been lucky or relied on its goalkeeper.
• npxG (non-penalty xG) — xG without penalties. A penalty is a high-value chance that depends on the referee more than on open play, so npxG shows more fairly how many chances a team creates by itself.
• xGOT (expected goals on target) — a rating of a shot after it was struck, counting only shots on target and taking into account where the ball was heading in the goal. A shot into the top corner gets a higher xGOT than a shot straight at the goalkeeper from the same spot. The gap between xGOT and goals conceded is often used to rate goalkeepers.
• xPts (expected points) — how many points a team would earn on average if its matches were replayed many times with the same xG. A table built on xPts shows who sits higher or lower in the real table than their chances justify. For an example from the current season, see our xPts team ranking.
• xG difference (xGD) — xG minus xGA. Over a long run it is one of the simplest summaries of a team's real level.

Can you calculate xG yourself?

A real xG model needs data for every shot: coordinates, body part, type of assist, position of defenders. That is why "xG calculators" that ask you to type in a number of shots only give a rough guess: without the location and type of each chance they cannot tell a tap-in from a long-range effort. Our xG calculator works the other way round: you place every shot on the pitch, and it rates each one by distance and angle to goal, adds the shots up into each team's xG and shows the win, draw and loss chances those shots were worth. It is a good way to see how the location changes a chance. For real matches it is more useful to take ready-made xG statistics from match data and work with them: compare a team's xG and goals over 10–20 matches, look at xG and xGA at home and away separately, and check whether a run of results is backed by the quality of chances.

Why xG and the final score diverge

Football is a low-scoring sport, so randomness — a striker's finishing on the day, a goalkeeper's reflexes, a deflection — affects the scoreline more than in high-scoring sports like basketball. Over a single match, a team with lower xG can absolutely win through finishing or luck. Over a longer run — 10, 20, 30 matches — results tend to converge toward xG: teams that consistently create better chances win more often over time.

Where to find xG on xG.Football

Every finished match page has a "Match statistics" section showing "Expected goals (xG)" for both teams, broken down by full match and by half. Each league also has its own season xG table with the xG, xGA and xPts of every team. To compare xG across a whole league season, open the competition tables and season statistics.

Using xG in football predictions

xG is one of the signals that lets you judge a team's underlying quality independently of the current league table. A team on a poor run of results but with a consistently high xG often turns its form around soon after — and vice versa. The xG.Football prediction model follows similar logic, combining points and goals with team form across several splits. See exactly how that works in "How the xG.Football Prediction Model Works", and check the model's real track record on the prediction statistics page.

Quick recap

Ready-made outcome probabilities for upcoming games, with xG as one of their inputs, are collected in the football predictions section.