Expected Goals, shortened to xG, is the single most important statistic in modern football. If you understand nothing else about football analytics, understanding xG will improve your betting immediately.
Here's the core idea: not all shots are equal. A penalty is worth far more than a long-range effort. A header from six yards out is worth more than a rebound from 30 yards away. xG converts every shot a team takes into a probability of becoming a goal, then adds those probabilities together to show how many goals they should have scored.
Why xG Matters More Than You Think
Watch a match where a team dominates but loses 1-0. The commentator calls it "one of those nights" or blames bad luck. But what if that team shot 15 times with a combined xG of 2.8? They didn't actually underperform. They just got unlucky in one match.
Now imagine that team repeats this pattern across five matches: dominant play, high xG, but poor results. After five matches, luck regresses. The team's points tally doesn't match their underlying performance. This discrepancy is where edges exist for bettors.
How xG Is Calculated
xG models analyse thousands of historical shots to understand which factors predict whether a shot becomes a goal. The main factors are:
Shot location: A shot from the penalty spot has roughly 76% xG. A shot from 20 yards has roughly 8% xG. Distance matters most.
Shot type: Penalties are highest. Headers are typically lower. Shots after one touch are higher than those after multiple touches.
Defensive pressure: Was a defender closing down the shooter? Were multiple defenders nearby? Pressure reduces xG.
Shot angle: A central shot from 12 yards is easier to score than a tight-angle shot from 8 yards.
Goalkeeper position: Did the goalkeeper have time to set, or was it a rushed situation?
Different xG models (Opta, StatsBomb, Sports Reference) weight these factors slightly differently, leading to small variations. But the core approach is identical: convert each shot into a probability between 0 and 1, then sum them up.
xG for Individual Players vs Teams
When you see "Erling Haaland has 15 goals from 12.3 xG," this means his expected goals based on shot quality is 12.3, but he's scored 15. He's overperforming his xG by 2.7 goals, suggesting he's a clinical finisher.
Conversely, a striker with 8 goals from 11.2 xG is underperforming, potentially due to poor finishing or simple bad luck.
For bettors, this matters because individual player performance tends to regress. Strikers who heavily overperform xG often see goal output decline. Those underperforming usually improve.
For teams, xG more reliably predicts future performance than goals scored. A team outscoring their xG consistently is likely unsustainable.
xGA: Expected Goals Against
The defensive equivalent of xG is xGA (Expected Goals Against), which shows how many goals a team should have conceded based on the chances given away.
A team with 1.8 xGA who has conceded 1 goal has been lucky. One with 1.8 xGA who has conceded 3 goals has performed poorly defensively. Over time, goals conceded regress toward xGA.
Using xG in Betting
xG is most useful when comparing it to results. Ask these questions:
Is there a gap between xG and results? If a team is outscoring their xG significantly, they may be overperforming luck and regression is coming.
Are both teams aligned? If both teams have similar xG but vastly different results, the underperforming team often bounces back.
Is the xG gap widening? If one team consistently creates 1.5+ xG more than opponents, they'll eventually convert this into wins.
Does xG match the scoreline? A 3-0 win where the winner had 1.2 xG and the loser had 1.4 xG suggests the match was closer to a coin flip than the scoreline suggests.
xG Performance Across European Leagues
Different leagues have varying levels of shot quality and finishing efficiency. Here's how average xG compares to actual goals across the top European leagues:
| League | Avg xG per team per match | Avg goals per team per match | xG vs Goals difference |
|---|---|---|---|
| Bundesliga | 1.52 | 1.58 | +0.06 (overperforming) |
| Premier League | 1.38 | 1.42 | +0.04 |
| La Liga | 1.29 | 1.35 | +0.06 |
| Serie A | 1.24 | 1.28 | +0.04 |
| Ligue 1 | 1.31 | 1.38 | +0.07 |
Note: Based on 2023-24 season data across top European leagues. Figures rounded to 2 decimal places.
The Bundesliga shows the highest xG and goal output, reflecting its faster pace and more open style of play. Serie A has the lowest, reflecting more compact, defensive football. When betting across leagues, these baselines help you evaluate whether a team's performance is sustainable. A Premier League team averaging 1.6 xG is performing well. A Serie A team with the same figure is significantly outperforming.
xG Limitations
xG doesn't account for goalkeeper quality. A shot with 30% xG against Alisson might be 25% against a lower-tier goalkeeper. Models can't easily assess this.
xG also doesn't account for individual striker skill with precision. A world-class finisher will consistently overperform xG slightly, though the effect is smaller than many believe.
Very rare events (bicycle kicks, deflections turning shots in) might register as 5% xG but feel like flukes. Over a season, these balance out.
Different xG Models
StatsBomb xG is considered most rigorous but is premium-priced. Opta provides xG through multiple platforms and is widely used. Sports Reference/FBref offers free basic xG data.
The differences between them are minor for strategic purposes. A shot from 12 yards that's 40% xG in one model might be 42% in another. This doesn't meaningfully change analysis.
In Summary
- XG converts every shot into a probability of becoming a goal, then sums these probabilities to show how many goals a team should have scored.
- This simple concept has enormous power because it separates luck from skill.
- Teams that consistently outperform their xG regress.
- Those underperforming it improve.
- Using xG alongside actual results reveals where the market might be mispriced.
- Start by looking at the xG difference in matches.
- If one team outshot the other significantly in quality, even if the result was even, that team has a good chance of winning next time.
- This is one of the strongest edges available to bettors.
FAQs
What's considered high or low xG in football? For a team in one match, 0.8-1.2 xG is typical for an attacking performance. Over 1.5 xG suggests strong attacking play. Below 0.5 xG suggests limited opportunities. For a season, strong teams often average 1.5+ xG per match, whilst poor ones average below 1.0.
Why do some teams always outperform their xG? Some teams do genuinely have better strikers. But most overperformance is luck. Over enough matches, regression occurs. Watch for stretches where a team does this, as corrective losses often follow.
Is xG more reliable than the actual result? For predicting future results, yes. xG is more predictive than goals scored of how a team will perform next. But for a single match, the actual result is the actual result. Use xG to predict forward performance, not to rewrite history.
Can xG be over 3.0 in a match? Rarely, but yes. A team that takes 20+ shots can accumulate high xG. The 2021-22 Manchester City season included matches with xG over 3.0. These matches are unusual.
How quickly does xG stabilise? Individual match xG is single data points. But across 5-10 matches, xG becomes meaningful. Across a season, it's highly predictive of points and goals. Use short-term xG to spot anomalies, not as gospel.
Does xG matter in all football leagues? xG works across leagues because it's based on fundamental shot probability. The Premier League and League Two both follow these principles. Understanding league-specific context (pace, defending style) matters alongside xG.