Expected Goals (xG) Explained: How to Use xG in Football Betting

Expected goals (xG) has transformed football analysis. It measures what should have happened based on the quality of chances, not what actually happened — and that distinction is gold for bettors.

What xG Is

Every shot in football can be assigned a probability of resulting in a goal. This probability is based on historical data from tens of thousands of similar shots, considering factors like:

Distance from goal
Angle to goal
Body part used (foot, head)
Type of assist (cross, through ball, cutback)
Whether the shot followed a dribble

A penalty has an xG of approximately 0.76. A one-on-one from 8 yards might be 0.45. A 25-yard volley might be 0.04.

Why xG Matters for Betting

The key insight: actual goals fluctuate randomly in the short term, but xG reflects the underlying quality of chances. A team that creates 2.1 xG per match but only scores 1.2 goals is underperforming — and regression to the mean suggests their results will improve.

Practical Application: Finding Value

Step 1: Gather xG Data

Check a team's xG for and xG against over their last 10-15 matches. Free sources include Understat and FBref.

Step 2: Calculate the xG Difference

Compare actual goals to xG:

Team A scored 15 goals from 11.2 xG — overperforming by +3.8 goals
Team B scored 8 goals from 13.5 xG — underperforming by -5.5 goals

Step 3: Identify the Value Bet

If Team B faces a mid-table opponent and bookmaker odds reflect their poor actual results (8 goals) rather than their underlying performance (13.5 xG), you may have found value on Team B.

xG for Different Markets

xG applies beyond match result:

Over/Under goals: Teams with high combined xG for and against tend to produce high-scoring games
BTTS: Check xG conceded — a team with high xG against is likely to concede
Correct score: xG-based Poisson models produce more accurate scoreline probabilities than goals-based models

Limitations to Understand

The Bottom Line

xG gives you a clearer picture of team performance than the scoreboard alone. When bookmaker odds are based on actual results that diverge significantly from underlying xG performance, you have a systematic way to find value.

Frequently Asked Questions

What does xG actually measure?+

xG assigns a probability to every shot based on historical data about similar shots. A penalty has an xG of about 0.76, meaning 76% of penalties are scored. A shot from 30 yards might have an xG of 0.03. The sum of all shots in a match gives the team's total xG.

Where can I find xG data?+

Free xG data is available from sites like Understat, FBref, and Infogol. These cover major European leagues including the Premier League, La Liga, Bundesliga, Serie A, and Ligue 1. Some offer match-level and shot-level data.

How do I use xG for betting?+

Compare a team's actual goals to their xG over the last 10-15 matches. If a team has scored 18 goals but their xG is only 12, they are overperforming and likely to regress. If they have scored 8 goals from 14 xG, they are underperforming and may improve.

Is xG better than actual goals for predictions?+

Over small samples, yes. Actual goals include random variance — lucky deflections, goalkeeper errors, and brilliant individual efforts. xG strips this noise out and gives a more stable measure of underlying performance. Over large samples, actual goals and xG converge.

What are the limitations of xG?+

xG does not account for the quality of the specific shooter, defensive pressure at the moment of the shot, or situational context like the game state. It is a probabilistic model, not a perfect predictor. It works best as one input among several.