xG Explained: How Expected Goals Works and How to Use It in Betting

Expected Goals (xG) has moved from an academic concept to a metric cited by television commentators, club technical departments, and serious football bettors worldwide. Understanding what xG actually measures — and what it does not — is essential for using it correctly in betting analysis.

What xG Measures

xG measures the probability that a specific shot will result in a goal, based on historical data about shots taken from the same position, angle, game situation and technique. A penalty carries around 0.76 xG. A header from the centre of the box under pressure might carry 0.09 xG. When all shots in a match are added together for each team, the total gives each team's xG — a measure of the total quality of chances created and conceded, independent of whether those chances actually went in.

Why xG Is Useful for Bettors

The primary value is identifying the gap between a team's results and their underlying performance quality. Football contains significant short-run variance — good teams lose matches they dominate statistically. A team that loses four matches in a row but consistently generates 2.0+ xG per match and concedes below 1.0 xG is performing at a level that does not match its results. These teams are structurally underpriced in subsequent fixtures because bookmakers and public bettors weigh results more heavily than the underlying data supports.

Practical Applications

Identifying Value After Bad Results

The most consistent xG application is identifying strong teams that have suffered bad results on good underlying performance. After three or four consecutive defeats that don't match their xG data, these teams are typically underpriced in the following fixture. Backing these teams when xG suggests the results are outliers produces value over large samples.

Identifying Overperforming Teams

Teams winning matches on low xG and high opponent xG are performing above their underlying level. They are getting better goalkeeping, more fortunate with opposition finishing, or benefiting from performances that typically revert to the mean. Backing their opponents at slightly inflated prices reflects the genuine underlying probability better than the market does.

Limitations of xG

• Goalkeeper quality is not captured — elite goalkeepers consistently save shots the model assigns high xG
• Striker quality is not captured — elite finishers convert at rates exceeding model predictions
• Sample sizes matter — xG over a single match is noisy; over twenty matches it is highly reliable
• Not all xG models are equal — be consistent about your data source

Summary: xG Application Principles

• xG measures chance quality, not outcomes — use it to identify performance divergence from results
• Teams with good xG and bad results are typically underpriced in subsequent fixtures
• Teams with bad xG and good results are typically overpriced — value exists on their opponents
• Goalkeeper and striker quality create legitimate structural xG divergence — not all outperformance is luck
• Use at least ten matches of xG data before drawing conclusions