What Is xG?
xG stands for Expected Goals. It’s a football statistic that measures the quality of a scoring chance by calculating how likely a shot is to result in a goal.
Every shot in a match gets an xG value between 0 and 1:
- xG of 0.01 = a very unlikely goal (e.g., a long-range shot from 35 yards)
- xG of 0.50 = a coin-flip chance (e.g., a one-on-one with the goalkeeper)
- xG of 0.90+ = a near-certain goal (e.g., a tap-in from two yards out)
How Is xG Calculated?
Statistical models analyse thousands — sometimes millions — of past shots to build xG formulas. They consider factors like:
- Distance from goal — closer shots have higher xG
- Angle to goal — central shots are easier than tight angles
- Body part — feet vs. head vs. other
- Type of assist — through ball, cross, set piece, etc.
- Game situation — open play, counter-attack, penalty
Different providers (Opta, StatsBomb, FBref) use slightly different models, so xG numbers can vary between sources.
A Simple Example
Imagine a team has three shots in a match:
| Shot | Description | xG |
|---|---|---|
| 1 | Long-range strike from 30 yards | 0.03 |
| 2 | Header from a corner, six yards out | 0.35 |
| 3 | One-on-one after a through ball | 0.55 |
That team’s total xG for the match would be 0.93 (0.03 + 0.35 + 0.55). On average, you’d expect them to score roughly one goal from those three chances.
Why Does xG Matter?
Measuring Team Performance
A team that consistently creates 2.0+ xG per match is attacking well, even if they’re not always scoring. Conversely, a team scoring from low-xG chances might be riding luck that won’t last.
Evaluating Players
A striker who regularly outperforms their xG — scoring more goals than expected — is considered a clinical finisher. One who underperforms xG may be wasteful in front of goal.
Predicting Results
xG tells a deeper story than the final score. A team that wins 1-0 but had 0.2 xG while the opponent had 2.5 xG was probably lucky. Over time, xG tends to be a better predictor of future results than actual goals scored.
Common xG Terms
- xG per 90 — xG generated per 90 minutes of play
- npxG — Non-penalty expected goals (excludes penalties, which are roughly 0.76 xG each)
- xG overperformance — when a player scores more than their xG suggests
- xGA — Expected Goals Against (how many goals a team is expected to concede)
What xG Does NOT Tell You
xG has limits:
- It doesn’t account for shot quality after deflections or lucky bounces
- It treats all shots the same regardless of the player taking them
- It doesn’t capture build-up play, pressing, or defensive actions
- Different models give slightly different numbers for the same shot
xG is best used alongside other stats and the eye test — not as the sole measure of a team or player.
xG vs. Actual Goals
Over one match, xG and actual goals can differ wildly. Over a full season, they tend to converge. That’s what makes xG so useful — it strips out randomness and reveals underlying performance trends.
Frequently Asked Questions
What is a good xG per match for a team?
An xG of around 1.5 to 2.0 per match is solid for most teams. Title contenders often average above 2.0. Anything below 1.0 suggests the team struggles to create chances.
Who invented xG?
The concept was popularised in the early 2010s by football analytics researchers. Sam Green and Marek Kwiatkowski at Opta, as well as public analysts like Michael Caley and StatsBomb’s Ted Knutson, helped bring xG into mainstream football discussion.
Is xG used by professional clubs?
Yes. Nearly every top-level professional club now employs data analysts who use xG and related metrics for tactical analysis, recruitment, and match preparation.
Can xG predict match outcomes?
xG-based models are among the most accurate public tools for predicting football results. However, football remains unpredictable — xG deals in probabilities, not certainties.