Understanding Expected Goals (xG) – The Ultimate Guide

Published: July 31, 2026 • 10 min read

Football Analytics

Expected Goals, commonly known as xG, has become one of the most widely used statistics in modern football analysis.

Traditional statistics tell us what happened. A team scored two goals, produced ten shots or had 60% possession. xG attempts to answer a different question:

How good were the chances that the team created?

This distinction is important because football scores can sometimes be misleading. A team can win a match despite creating fewer and weaker chances, while another team can lose after producing several excellent opportunities.

xG provides a way to examine the quality of those chances and understand performances beyond the final score.

What Is Expected Goals?

Expected Goals is a statistical measure that estimates the probability of a particular shot becoming a goal.

Each scoring opportunity receives an xG value between 0 and 1.

For example:

• A shot with an xG of 0.05 has an estimated 5% chance of becoming a goal.
• A chance with an xG of 0.25 has an estimated 25% chance.
• A chance with an xG of 0.80 has an estimated 80% chance.

These numbers are not predictions of what will happen to an individual shot.

A chance with 0.80 xG can still be missed.

Instead, xG is based on what happened to a large number of similar chances in the historical data used to build the model.

How Is xG Calculated?

Different statistical providers use different models, but xG systems generally consider information about the circumstances surrounding a shot.

Depending on the model, factors can include:

• Location of the shot
• Distance from the goal
• Angle to the goal
• Body part used
• Type of assist
• Whether the chance followed a cross
• Whether it followed a through ball
• Whether the shot came from a set piece
• Whether the opportunity followed a rebound
• Defensive pressure
• Previous actions in the attacking sequence

The more information a model incorporates, the more detailed its estimate can become.

This is why xG values can differ between data providers. There is no single universal xG number for every chance.

A Simple Example

Imagine a team produces five shots:

The combined xG is:

0.03 + 0.08 + 0.18 + 0.45 + 0.60 = 1.34 xG

This means the team created chances that, according to the model, would be expected to produce approximately 1.34 goals over a large number of similar situations.

It does not mean the team is guaranteed to score one or two goals in that particular match.

It could score zero, one, three or more.

Why xG Is Useful

Goals are relatively rare events.

A team can score from an unlikely chance and make its performance appear better than it actually was.

For example, suppose Team A wins 1–0 after scoring from a long-range shot.

Meanwhile, Team B creates several close-range opportunities but fails to score.

The final score says:

Team A 1–0 Team B

But the chance quality might look very different:

Team A: 0.45 xG

Team B: 1.80 xG

The result belongs to Team A, but Team B may have produced the more dangerous attacking performance.

This is one of the main reasons analysts use xG alongside traditional results.

xG For and xG Against

There are two basic sides to team xG analysis.

xG For

xG For measures the quality of chances a team creates.

A team regularly producing high xG is generally creating more dangerous scoring opportunities.

xG Against

xG Against, often abbreviated as xGA, measures the quality of chances a team allows its opponents to create.

A team consistently recording low xGA is generally restricting opponents to lower-quality opportunities.

Looking at both numbers gives a more complete picture.

For example:

Team A is creating considerably more quality chances while allowing fewer.

That does not guarantee a victory in the next match, but it provides useful evidence about the teams' underlying performances.

Understanding xG Difference

Another useful metric is xG difference.

The basic calculation is:

xG Difference = xG For − xG Against

Suppose a team averages:

1.70 xG

and:

0.95 xGA

Its xG difference is:

+0.75

A positive number indicates that the team is creating more expected goals than it is allowing.

A negative number indicates the opposite.

Over a larger sample, xG difference can be useful for comparing overall team strength.

xG Per Match

Average xG can help compare attacking performance between teams.

For example:

Team A: 1.90 xG per match

Team B: 1.15 xG per match

Team A has generally been creating better-quality chances.

However, the quality of opponents matters.

A team playing against weaker defenses may naturally produce higher xG numbers than a team facing elite opponents every week.

Always consider the competition and opposition when interpreting averages.

xG and Actual Goals

One of the most interesting uses of xG is comparing expected goals with actual goals.

Suppose a team has:

20 goals scored

from:

15 xG

The team has scored five more goals than its expected total.

That could be caused by several factors, including excellent finishing, particularly effective players, goalkeeper errors, chance-quality differences not captured by the model or simple statistical variation.

Similarly, a team with:

10 goals

from:

17 xG

has scored significantly fewer goals than its chance quality would suggest.

This does not automatically mean that the team is "due" to score more goals.

However, it can indicate an area worth investigating.

Overperformance and Underperformance

The difference between goals and xG is sometimes used to identify finishing overperformance or underperformance.

Example

A striker has scored:

12 goals from 8 xG

This represents substantial overperformance relative to the model.

Another striker has scored:

5 goals from 9 xG

This represents underperformance.

But it is important not to assume that all overperformance will immediately disappear.

Elite players can consistently outperform certain xG models because of finishing ability, shot selection or other characteristics not fully captured by the model.

Therefore, xG should identify questions rather than automatically provide answers.

What Does xG Say About Finishing?

xG is particularly useful when evaluating whether goals are being supported by chance creation.

Imagine two teams:

Team A

• 1.90 xG
• 2.00 goals per match

Team B

• 0.95 xG
• 1.80 goals per match

Both teams are scoring frequently.

But Team A's goal production is more closely aligned with its chance creation.

Team B is scoring substantially more goals than its average chance quality might suggest.

That does not prove Team B will decline, but it is a reason to investigate further.

Shot Quantity vs Shot Quality

One of the biggest advantages of xG is that it distinguishes between how many shots a team takes and how dangerous those shots are.

Consider two teams.

Team A

• 18 shots
• 4 shots on target
• 0.75 xG

Team B

• 9 shots
• 5 shots on target
• 1.70 xG

Team A took twice as many shots.

But Team B created much better opportunities.

This demonstrates why raw shot totals can be misleading.

More shots do not automatically mean better attacking performance.

xG and Possession

Possession is another statistic that becomes more useful when combined with xG.

A team can dominate possession without creating dangerous chances.

For example:

65% possession

but:

0.70 xG

This could indicate that the team controlled the ball without creating enough penetration.

Another team might have:

35% possession

and:

1.60 xG

This could indicate a highly effective counter-attacking approach.

The important question is not simply how much of the ball a team has.

It is what the team does with that possession.

xG and Home Advantage

Home and away xG can reveal information that overall averages hide.

A team might have:

Home xG: 1.90

Away xG: 1.05

That difference suggests its attacking performance changes significantly depending on venue.

The same analysis can be applied to xGA.

For example:

Home xGA: 0.85

Away xGA: 1.45

This could indicate that the team is considerably more defensively reliable at home.

When analyzing a specific fixture, home and away xG can therefore be more relevant than a single season-long average.

xG in Match Analysis

xG can help explain individual matches.

Imagine a match finishes:

Team A 0–1 Team B

The basic result suggests that Team B was better.

But suppose the statistics show:

Team A: 2.10 xG

Team B: 0.65 xG

Team B won the match, but Team A created considerably better opportunities.

This does not mean the result was "wrong."

Football is decided by actual goals, not expected goals.

Instead, xG provides additional context about how the match developed.

xG in Betting Analysis

For people analyzing football betting markets, xG can be particularly useful as one component of a broader process.

It can help answer questions such as:

• Is a team's recent scoring record supported by its chance creation?
• Is a defense allowing high-quality opportunities?
• Is an attacking improvement visible in the underlying numbers?
• Are recent results consistent with performance?
• How do two teams compare in chance creation?

However, xG should never be treated as a guaranteed prediction system.

A team with a strong xG profile can still lose.

A team with poor xG numbers can still win.

The value comes from combining xG with other information.

xG for Over/Under Analysis

Goal markets are another area where xG can provide useful context.

Suppose two teams have recently produced:

Team A: 1.75 xG

Team B: 1.65 xG

Both teams are also allowing relatively high-quality chances.

That profile may suggest a match with considerable attacking potential.

But the analysis should also consider:

• Recent goal totals
• xGA
• Home and away splits
• Injuries
• Expected lineups
• Tactical styles
• Rest days
• Opponent quality

A single xG number should never be used as an automatic signal for an over or under selection.

xG for Both Teams to Score Analysis

BTTS markets ask whether both teams will score.

xG can help evaluate the attacking capability of both sides.

For example, if both teams consistently create good chances and neither defense has been particularly effective at limiting opponent opportunities, that may provide evidence worth investigating.

However, the relevant question is not simply:

"Do both teams have high xG?"

You should also consider whether each team's chances are likely to translate into goals against this particular opponent.

xG Does Not Predict Individual Matches Perfectly

This is one of the most important things to understand.

Suppose a team has:

2.00 xG

in a match.

It does not mean the team will score exactly two goals.

The team could score:

• 0 goals
• 1 goal
• 2 goals
• 3 goals
• 4 or more goals

The expected value represents an average across many comparable situations.

Football contains considerable randomness, which is why even strong statistical models cannot eliminate uncertainty.

Different xG Models Can Produce Different Numbers

There is no universal xG model.

Different providers may use different:

• Data sources
• Variables
• Historical samples
• Methodologies
• Treatment of penalties
• Treatment of set pieces

As a result, one provider might assign a chance an xG of 0.35 while another assigns 0.29.

This does not necessarily mean one model is wrong.

When tracking xG over time, consistency is important. Comparing numbers from the same provider can make trends easier to interpret.

Penalties and xG

Penalties are a special case in expected-goals models.

A penalty is a relatively high-probability scoring opportunity compared with most open-play shots.

Depending on the model, penalties may receive a high xG value or may be treated separately.

This matters when comparing teams.

A team that has received several penalties can have a significantly higher xG total than its open-play chance creation alone would suggest.

Therefore, when performing detailed analysis, it can be useful to understand how the data provider treats penalties.

Set Pieces and xG

Corners and free kicks can also contribute significantly to a team's expected goals.

Some teams have highly effective set-piece routines and consistently create dangerous opportunities from dead-ball situations.

Others may generate most of their xG through open play.

When comparing teams, look beyond the total number and consider where the expected goals are coming from.

This can reveal tactical strengths that basic statistics may miss.

Common xG Mistakes

Mistake 1: Treating xG as the Final Score

xG does not replace actual goals.

The scoreboard determines the result.

xG provides additional analytical context.

Mistake 2: Assuming High xG Guarantees a Win

A team can create 2.5 xG and still lose.

Mistake 3: Looking at Only One Match

One match is usually too small a sample for meaningful conclusions about long-term team quality.

Mistake 4: Ignoring Opponent Quality

Producing 2.0 xG against a weak defense is different from producing 2.0 xG against an elite defense.

Mistake 5: Assuming Regression Is Guaranteed

Underperforming xG does not guarantee that goals will suddenly increase.

Mistake 6: Comparing Different Models Without Context

Different providers can calculate xG differently.

Mistake 7: Ignoring Team News

A team's previous xG numbers may not accurately represent its current lineup.

How to Use xG Properly

A practical xG analysis can follow this process:

Step 1: Look at the sample size

Use multiple matches rather than relying on one result.

Step 2: Check xG For

Evaluate how many quality chances the team creates.

Step 3: Check xGA

Evaluate the quality of chances the team allows.

Step 4: Calculate the xG Difference

Compare attacking production with defensive exposure.

Step 5: Compare Home and Away Numbers

Look for venue-specific differences.

Step 6: Compare xG With Actual Goals

Identify potential differences between chance creation and finishing.

Step 7: Examine the Opponents

Consider the quality of the teams faced.

Step 8: Check Team News

Adjust your interpretation if important players are unavailable.

Step 9: Consider Tactical Matchups

Determine whether the opponent's style could change the expected pattern.

Step 10: Combine the Evidence

Use xG as one part of a broader football analysis.

A Practical Example

Imagine two teams preparing for a league match.

Team A

• 1.80 xG per match
• 0.95 xGA
• 2.0 goals scored
• 0.9 goals conceded
• Strong home record

Team B

• 1.15 xG per match
• 1.50 xGA
• 1.2 goals scored
• 1.6 goals conceded
• Weak away record

On the surface, Team A appears to have the stronger statistical profile.

But a complete analysis should still ask:

• Who did each team face?
• Are the numbers based on enough matches?
• Are important players missing?
• Has either team recently changed manager?
• Does Team B's tactical style create problems for Team A?
• Are the home and away numbers consistent with the overall figures?

The statistics provide a starting point, not a guaranteed conclusion.

Building a Better Football Model With xG

xG becomes more powerful when combined with other indicators.

A broader model might include:

Attacking

• xG
• Goals
• Shots
• Shots on target
• Big chances

Defensive

• xGA
• Goals conceded
• Shots allowed
• Clean sheets

Context

• Home/away performance
• Recent form
• Opponent strength
• Injuries
• Suspensions
• Rest days
• Tactical matchup

This approach creates a much more complete picture than relying on xG alone.

Final Thoughts

Expected Goals has changed the way many analysts evaluate football.

Instead of looking only at the final score, xG allows us to examine the quality of chances created and conceded.

It can help explain why a team won despite being outplayed, why another team lost despite creating excellent opportunities, and whether recent scoring records are supported by underlying chance creation.

But xG is not a crystal ball.

It is a statistical model designed to provide additional information about scoring opportunities.

The most effective way to use it is to combine xG with actual results, defensive statistics, home and away performance, team news, tactical analysis and the quality of opposition.

The key principle is simple:

Goals tell you what happened. xG helps you understand the quality of the chances behind what happened.

Once that distinction is understood, football statistics become much more useful—and much harder to misinterpret.