A football game’s full-time result doesn’t always reflect the number and quality of chances created throughout the 90 minutes. xG (expected goals) is an underlying metric which provides additional context beyond the full-time result.
But what is xG in football stats? Every shot a player takes is assigned a number between zero and one for its likelihood of resulting in a goal.
The shot’s distance from goal, angle, and assist type all impact the xG figure. For example, a long-range effort may have an xG of 0.07, while a penalty kick may be valued at 0.78.
xG is used by professional football clubs, sportsbooks, and bettors alike to understand the underlying patterns and trends of a team or player.
Key findings
- xG gives every shot a value between zero and one, based on how often similar efforts have resulted in goals.
- A team’s total xG is the sum of all its shots’ values, showing how many goals its chances were worth rather than predicting the result.
- One match’s xG rarely tells the full story, but expected goal difference is a strong predictor of points over a season.
- Penalties are valued between 0.75 and 0.79 depending on the model, which is why analysts often use non-penalty xG (npxG).
- Bettors can use xG to find value where a player’s recent goal record and underlying chances do not match, but sportsbooks price markets using many more variables.
xG explained: What does xG mean in football?
In short, the xG football meaning is a shot’s probability of resulting in a goal, based on the conversion rates of previous efforts taken under similar circumstances.
For example, a shot from an unfavourable angle inside the penalty box might have an xG of 0.30, meaning that it would be expected to be converted on three out of ten occasions.
A team’s xG is calculated by adding together the xG values of all its shots during a match.
Total xG is often misinterpreted as a prediction of how many goals a team will score over 90 minutes.
Instead, it’s the total of each of their shots’ xG, showing how many goals their chances were worth.
During a 90-minute game, a team may take seven shots with the following xG values: 0.03, 0.21, 0.16, 0.71, 0.38, 0.08, 0.18, meaning their total xG is 1.75.
| Chance | Real-World Example | Example xG | Meaning |
| Difficult long-range shot | Kylian Mbappe’s long-range effort against Croatia in the 2018 World Cup final. | 0.03 | 3% chance |
| Reasonable opportunity | Mario Gotze’s winning goal in the 2014 World Cup final against Argentina. | 0.38 | 38% chance |
| Very strong opportunity | Lionel Messi’s close-range goal against France in the 2022 World Cup final. | 0.86 | 86% chance |
How is xG calculated in football?
Football fans may watch a goal-scoring opportunity in real time and believe that it’s a gilt-edged chance, only for an xG model to assign it a low probability.
So, how is football xG calculated?
All xG models use criteria and algorithmic data which predict the likelihood of a shot resulting in a goal.
xG data models consider factors such as distance to goal, angle to goal, body part used, and type of assist or previous action, using the outcomes of previous efforts with similar characteristics.
For example, two shots taken with the foot from the same position could have different xG figures if one pass was weighted poorly and received from an unfavourable angle.
While xG figures don’t vary massively between different data model providers, different criteria can be used to calculate a shot’s probability of resulting in a goal.
Hudlstatsbomb’s model evaluates additional information, such as the goalkeeper’s positioning and the space in front of the attacker.
A shot assigned an xG of 0.32 by one data provider may be recorded as 0.37 by another.
xG Meaning: What is a good xG in football?
How does xG work in performance analysis?
There is no definitive good or bad xG figure for an individual match, as the number of chances created and the game state can significantly influence a team’s total.
In Liverpool’s Champions League finals against Real Madrid in 2018 and 2022, the Premier League club accumulated a combined total xG of 3.50 to Los Blancos’ 2.29, but lost both matches by an aggregate score of 4-1.
Over one or two matches, such as Liverpool’s finals against Real Madrid, xG may not tell the full story.
While Real Madrid only accumulated an xG of 0.92 in the 2022 final, they scored in the 59th minute and defended their lead.
By contrast, Liverpool attacked Madrid’s goal to equalise, eventually ending the contest with an xG of 2.19.
xG figures can’t be perceived as good or bad from one match without the game’s context or the team’s tactics.
However, teams that consistently create more xG tend to score more goals over time. xGD (expected goal difference) is a team’s xG minus xGA (expected goals against).
In xGA football statistics, xGA measures the quality of the chances a team allows its opponents to create.
A 2025 study found that xGD was one of the strongest predictors of team success, with a correlation of 0.91 with points per match in 2023-24 and 0.93 in 2024-25.
xG vs actual goals: What does the difference tell us?
A player who underperforms their xG scores fewer goals than expected based on the quality and quantity of their chances.
A player’s goals can fall below their xG because of short-term bad luck, such as goalkeepers making unusually good saves, defenders making last-ditch blocks or shots hitting the woodwork.
However, persistent underperformance can also indicate poor finishing.
Recruitment teams may identify players who are underperforming their xG and potentially spot value in attacking players that have been unlucky in front of goal.
However, xG underperformance alone cannot explain whether a player had bad luck or poor finishing skills.
Other players may overperform their xG. During the first three weeks of the 2026-27 season, seven Premier League players overperformed their xG by one goal or more.
xG over- or underperformance is calculated by subtracting xG from actual goals.
For example, Anthony Elanga scored two goals from an xG of 0.42, meaning his xG overperformance is 1.58.
The metric is used by sportsbooks and bettors, too.
A strategic bettor may therefore be cautious about backing a player to score if they have been significantly overperforming their xG, particularly if their underlying chance creation has remained unchanged.
However, elite players can consistently overperform their xG. For example, during the 2025-26 Bundesliga, Harry Kane overperformed his xG by 7.59 goals.
What is the xG of a penalty?
Each data model values a penalty’s xG differently, but it typically falls between 0.75 and 0.79.
Opta Analyst’s penalty xG is 0.79; Hudl StatsBomb values spot-kicks at 0.78, while most other traditional models value them at 0.76.
Over the last nine Premier League seasons, penalty conversion rates have fluctuated from 70.0%–89.7% by year.
With a value of around 0.75–0.79, penalties can significantly influence a player’s total xG value, making their performances harder to interpret.
npxG (non-penalty expected goals) subtracts penalties’ xG value from totals.
This model makes sense for recruitment teams, since penalties can inflate a player’s xG total without necessarily reflecting their ability to create high-quality chances through intelligent movement and positioning.
However, bettors who are analysing the stats of a team’s designated penalty taker may choose to use xG instead.
How accurate is xG?
Some media pundits criticise xG without fully understanding its purpose.
A team that accumulates a higher or lower total xG doesn’t automatically deserve to win or lose.
In Liverpool’s 2-0 win over Ipswich, they accumulated an xG of just 0.66 compared to the opposition’s 0.73.
However, Liverpool were two goals ahead by the ninth minute and didn’t need to continue attacking to win the game.
Likewise, a team that underperforms xG in several consecutive matches isn’t ‘due’ a winning streak. xG underperformance can be influenced by several factors, including poor finishing.
xG explains how many, and of what quality, chances a team creates, but it doesn’t factor in the quality of an attacking player.
For example, if Erling Haaland and an 18-year-old debutant both take a penalty, the xG value remains the same.
Given that goal-scoring and clinical finishing are skills in themselves, it can be argued that a team that wins a game by creating an xG of 0.24 is no less deserving than one that accumulates 3.25.
Therefore, xG measures chance creation rather than win-deservedness.
xG stats may also not offer insights about future outcomes if attacking players tasked with scoring goals are significantly stronger or weaker than the average.
How is xG used in football betting?
xG and related expected-performance metrics can provide valuable information for football betting analysts.
- By analysing xG data, bettors can assess a team’s attacking strength and defensive performance beyond its recent results.
- xG can also help bettors estimate goal expectation and expected goals prediction for teams and players.
For example, a bettor may notice that a player who has a historic record of matching their xG has been underperforming it in recent weeks.
On some occasions, that may make betting odds on ‘X player to score’ markets increase.
While sportsbooks use algorithms and consider underlying data metrics when pricing markets, if the public underestimates a player due to their recent scoring record, the player’s goal-scoring odds may increase.
However, sportsbooks and betting data providers do not simply take publicly available xG football figures and convert them directly into odds.
Sophisticated pricing models can incorporate numerous variables alongside expected-goal-style metrics.
Team strength, expected and confirmed line-ups, injuries and suspensions, and home advantage, among several other factors, are used to price betting markets.
How to read xG in a football match
Consider a hypothetical match between Barcelona and Arsenal:
- Final score: Barcelona 1-0 Arsenal
- xG: Barcelona 0.7 – Arsenal 1.8
Despite losing the match, Arsenal generated chances that collectively carried a higher xG value than Barcelona’s chances.
However, Barcelona scored the game’s only goal and won.
Whether Barcelona or Arsenal deserved to win the game is open to interpretation and subjective.
This demonstrates the fundamental point:
xG describes chance quality, but it does not determine the score.
It also explains why analysts can use xG to look beyond individual results when assessing underlying performance.
