Size = xG
xG (Expected Goals) is the best single number for measuring the quality of chances a football team creates. This guide walks you through how xG is calculated, how to read a shot map, and why xG predicts results better than the scoreline.
The one-sentence answer
xG is the probability that a given shot will be scored, based on where it was taken from, how it was taken, and the defensive situation around it. Added up across a match, xG tells you the quality of chances each team created, not just who finished them.
Every shot is fed into a model trained on hundreds of thousands of historical shots. The model weighs six big factors and spits out a probability between 0 and 1.
Shots closer to goal are more likely to be scored. Distance is the single biggest factor in any xG model.
A shot from the centre of the box has a wider target than a tight angle from the byline. Angle captures how much of the goal the shooter can actually aim at.
Feet, head, weaker foot. Headers and weak-foot shots convert at lower rates than strong-foot shots from the same position.
Open play, set piece, free kick, penalty. A penalty is worth ~0.76 xG regardless of who takes it; other shot types follow their own base rates.
Through balls, crosses, cut-backs and rebounds all change the defensive context. A cut-back to the penalty spot is more dangerous than a cross at the same distance.
Where defenders stand, how many, and whether the goalkeeper is set. Pressure deflates xG; space inflates it.
xG Stat runs on independent professional event data. Different providers weight these factors slightly differently, which is why xG values for the same shot can vary across Opta, StatsBomb, and Wyscout by up to ~0.1.
Every dot is a shot. Bigger dots mean higher xG. A match with lots of big dots but no goals tells a different story from one with one lucky long-ranger.
Big dot in the middle of the six-yard box = a near-certain goal (0.6+ xG). If a team creates several of these and doesn't score, expect regression.
Small dots outside the box = low-quality chances (under 0.05 xG). Lots of these usually means the team couldn't break through a packed defence.
Clusters of dots in one zone = tactical signal. A team finding the same attacking area repeatedly is a plan working, not luck.
Goals are rare and random. xG is the signal underneath them. That's why it's now part of the language of football.
The xG standings show where teams should sit based on the chances they create and concede. It cuts through lucky wins and unlucky defeats.
See the xG standingsStrikers who outperform their xG over a few matches are often just lucky. Over a full season, xG filters finishing streaks from genuine production.
Browse the player databasexG is the single best leading indicator for future goals. Captaincy, transfers, and match previews all get sharper when you use it.
Open the FPL toolsThe scoreboard tells you who won. xG tells you who should have. Over a single match they can diverge wildly; over a season they almost always converge.
Every step up on the line is a shot, sized by its xG. Flat stretches are drought periods. The xG timeline shows how a match actually felt minute by minute, which a 1-0 scoreline can hide completely.
Myth
Reality
xG isolates chance quality so you can measure finishing. If a striker consistently outperforms their xG over 100+ shots, that's a real finishing signal.
Myth
Reality
Missing a 0.8 xG chance feels bad, but getting into a 0.8 xG position is the skill. Shot selection matters more than any single miss.
Myth
Reality
Betting markets use xG, yes, but so do coaches, scouts, fantasy managers, and analysts. It's a general-purpose quality-of-chance measure.
Myth
Reality
Modern xG models tag each shot with its build-up type. Set pieces and penalties have their own base rates baked into the number.
Rule of thumb: below 0.05 xG is a speculative shot, around 0.1 is a half-chance, 0.3+ is a clear chance, and 0.76 is a penalty. Anything above 0.5 is a shot you'd expect to be scored at least half the time.
No. Opta, StatsBomb, Wyscout, and Understat all train their own models on their own event data. Values for the same shot can differ by 0.05 to 0.1. xG Stat runs on independent professional event data.
Yes, and every model uses a similar base rate around 0.76 xG per penalty. Some charts exclude penalties to make open-play comparisons cleaner.
xGOT (expected goals on target) rates the quality of the shot on target, factoring where in the goal frame it was placed. It measures finishing and goalkeeper performance, where xG measures chance quality.
For individual players, no. xG is noisy over small samples. For team-level match analysis it's already useful: a team with 2.5 xG probably played better than one with 0.4 xG, regardless of the scoreline.
Different providers track slightly different events, define shot types differently, and train on different historical datasets. xG Stat runs on independent professional event data, so season-level trends agree even where the decimals do not. For FBref-style coverage, see our FBref alternative guide.
Every Premier League match. Shot maps, timelines, pass networks, and the story behind the scoreline.
Sign Up