xG Guide: expected goals explained
Everything about expected goals, from the first idea to the fine print: what xG is, how it is measured, who measures it, what it can and cannot tell you. Sourced, dated and free, from the livescore that shows xG live.
The xG family
Part 1 · The basics
What xG is, how it is made, and how to read it right.
Expected goals, xG for short, is the chance that a shot becomes a goal, written as a number from 0 to 1.
Chapter 2How xG is calculatedAn xG model looks at a shot the moment before it is struck, compares it with a huge archive of past shots, and returns the share that ended in a goal.
Chapter 3Typical xG valuesWhat is a penalty worth? A header? A shot from 25 metres? Here are the figures that published sources support, each with how firm it is.
Chapter 4Why providers disagreeThere is no official xG.
Chapter 5Live xGLive xG is xG built while the match is played: shot by shot, from data notated in real time.
Chapter 6xG mythsxG is easy to quote and easy to misread.
Chapter 7Shot-value explorerPlace a shot and watch its value change.
Chapter 8xG calculatorTwo numbers, every chance.
Part 2 · Reading xG
What xG can tell you about matches, teams and players, and how sure it is.
Two pictures tell a match's xG story: the shot map shows where the chances were and how good they were; the xG race shows who created them, and when..
Chapter 10One big chance or many smallTwo teams finish with 0.60 xG.
Chapter 11xG for, against and differenceA team's xG says how good its chances were.
Chapter 12Expected pointsReplay a match's chances many times, with each shot going in at its own xG.
Chapter 13Sample sizeOne match is a coin toss with a few shots.
Chapter 14Hit rate and base rate"Right 50% of the time" sounds like something.
Chapter 15Finishing over xGA striker scores 20 goals from 14 xG.
Chapter 16Game stateA team 2-0 up does not play like a team 0-1 down.
Chapter 17How accurate is xG?An xG model is a set of probabilities.
Part 3 · Beyond the shot
The metrics around xG: after the shot, before the shot and beyond shots.
xG values the chance before the ball is struck.
Chapter 19Goals preventedA goalkeeper can only stop what comes on target.
Chapter 20Expected assistsExpected assists credit the passer for the chance he creates, whether or not the shooter scores.
Chapter 21xGChain and xGBuildupA goal starts long before the shot.
Chapter 22Expected threat (xT)xG only sees shots.
Chapter 23OBV and VAEPThe next step after xG and xT: value every action on the ball, including tackles and clearances, by how it changes the chance of scoring and of conceding..
Chapter 24PPDA and field tiltTwo context numbers often shown next to xG: one for how hard a team presses, one for how much of the game it plays in the opponent's half..
Chapter 25Set piecesCorners, free kicks and penalties produce a large share of goals.
Chapter 26Women's football xGA shot from the same spot does not have the same chance of going in everywhere.
Part 4 · Context
Where xG came from, and what it is not for.
Expected goals did not appear overnight.
Chapter 28xG and oddsMany sites sell xG as a betting edge.
Reference
Every term in the expected goals family, in plain words.
ReferencexG FAQEvery question answered in the guide, in one place, each linked to the chapter that explains it in full..
ReferenceSourcesEvery figure in the guide comes from a named, dated source.
How this guide works
Read it in order, part by part, or open any chapter on its own: each one stands alone. Every figure comes from a named, dated source, listed on each page. Where providers disagree, we show both. Where nobody publishes a number, we say so rather than guess. And where Xgoals itself is involved, a lime box says exactly what we show and what we do not.
Questions
Is the xG Guide free?
Yes. No account, no paywall, no ads.
Where do the figures come from?
From the data companies' own explanations (Opta, Hudl StatsBomb, Wyscout and others) and from published research, each named and dated on its page.
xG explains football; it is not a betting tip. 18+ · Please gamble responsibly.