Premier Betting Tips • Betting Guides
Published 4 October 2026 · Educational guide
Quick answer: Expected goals, or xG, estimates the chance that a shot becomes a goal based on its characteristics. A shot valued at 0.20 xG represents an estimated 20% scoring chance under that model. Adding shot values summarises the chances a team created; it does not change the actual score.
A team can win despite creating fewer high-quality chances, or lose after producing several good opportunities. xG helps you examine that difference. It is useful evidence, but neither a verdict on who deserved to win nor a guaranteed forecast of the next match.
How is a shot’s xG calculated?
Models learn from historical shots. Typical inputs include distance, angle, body part and the situation preceding the shot. Richer datasets can include goalkeeper and defender positions. Providers use different data and modelling choices, so the same chance can receive different values.
A model assesses the situation; it does not claim that a player scores a fraction of a goal. The actual shot ends in a goal or no goal. A low-value shot can score and a high-value shot can miss.
A simple three-shot example
These invented values illustrate the arithmetic, rather than fixed values for every chance of each type.
| Chance | Example xG | Model interpretation |
|---|---|---|
| Long-range attempt | 0.05 | Estimated 5% chance of scoring. |
| Chance inside the penalty area | 0.25 | Estimated 25% chance of scoring. |
| Close-range opportunity | 0.60 | Estimated 60% chance of scoring. |
| Total | 0.90 | Sum of the three shot estimates. |
Team A could score none, one, two or all three attempts. The total 0.90 is the expected number of goals from those recorded shots under the model. It is not a 90% chance of winning, nor automatically a 90% chance of scoring at least once.
xG, xGA and xG difference
| Label | Meaning |
|---|---|
| xG / xG for | The modelled chance quality of shots a team takes. |
| xGA / xG against | The modelled chance quality of shots it concedes. |
| xG difference | xG for minus xG against over the same period. |
| Non-penalty xG | xG excluding penalty attempts, under the provider’s definitions. |
| xGOT / post-shot xG | A different assessment that incorporates shot execution, such as where an on-target attempt goes. |
If Team A records 1.80 xG and concedes 0.70 xG, its xG difference is +1.10. That is a chance-quality difference, not a predicted 1.10-goal winning margin or an Asian handicap recommendation.
Why can the score disagree with xG?
Chance conversion varies. Finishing, goalkeeping and random variation can lead to actual goals above or below the expected total. A single match is particularly vulnerable to a small number of decisive events.
Imagine Team A loses 0–1 despite xG totals of 1.80–0.70. The result remains a defeat. The chance record suggests questions worth investigating: what kinds of chances occurred, when did they happen, and how did the score affect the teams’ approach?
Do not turn that example into a claim that Team A will win next time. Performance can change with opponents, personnel and tactics. A gap between goals and xG does not guarantee a correction in the next fixture.
How can xG help assess goal markets?
Comparable xG for and against can contribute to an assessment of attacking and defensive strength. Look at venue, opponent quality, sample size and whether the figures include penalties. A post-match total describes shots already taken; a pre-match forecast requires a separate model of the upcoming fixture.
Do not assume a historical combined average of 2.60 xG guarantees over 2.5 goals. An average does not give the whole distribution of possible totals. Read the over 1.5 versus over 2.5 guide to understand the thresholds.
Why full-match xG cannot replace first-half xG
For a first-half goal assessment, the period matters. A team might create most of its chances after the interval. Full-match figures do not tell you which half produced them.
Dividing a full-match figure by two assumes chances are distributed evenly. Unless a validated model justifies that assumption, it is not a measured first-half statistic. If the strategy requires first-half xG and that input is missing, record the limitation.
The first-half goals guide explains how to read provisional assessments and NO SCORE labels.
How does xG relate to Asian handicaps?
A forecast may use attacking and defensive estimates to model scorelines and winning margins. Handicap value then depends on the exact line, odds and probabilities of each settlement outcome.
A positive xG difference alone does not tell you whether −0.5, −1 or −1.25 is suitable. These lines settle differently. Use the Asian handicap settlement guide alongside the analysis.
xG versus FIS, confidence and probability
- xG: A shot-quality metric, or an aggregated total of those estimates.
- FIS: A rating under the report’s stated assessment framework.
- Confidence: A separately defined measure, such as evidence reliability.
- Outcome probability: A forecast for a specified event, such as over 2.5 goals.
These values cannot be substituted for one another. Entering xG 1.80 as a probability makes no sense; nor does entering FIS 80 as 80% without a justified probability forecast. See the FIS and confidence guide and calculator instructions.
Five checks before relying on xG
- Identify the provider. Avoid combining incompatible model values without explanation.
- Check the scope. Full match, first half, team total and non-penalty figures describe different things.
- Review the sample. A few matches against unusual opposition may give a distorted impression.
- Examine context. Red cards, penalties, score effects and lineup changes can alter the interpretation.
- Separate analysis from price. Useful performance evidence does not prove a bet is good value at current odds.
Frequently asked questions
Does 2.0 xG mean the team should definitely score two?
No. It is an expected total under the model for the recorded shots, not a guaranteed integer result.
Can xG be higher than the actual goals?
Yes. A team can miss chances with substantial estimated scoring probabilities.
Is xG the same as shots on target?
No. Shot counts measure quantity. xG estimates chance quality; an off-target attempt can still have pre-shot xG.
Is xGOT the same as ordinary xG?
No. xGOT adds information about on-target shot execution. Check the provider’s definitions before comparing the metrics.
Sources
Definition and shot modelling: Hudl Statsbomb: Expected Goals explained. Model features and limitations: Hudl Statsbomb: Upgrading Expected Goals. Post-shot distinction: Stats Perform: Introducing xGOT. Team-level interpretation: Statsbomb: Analysing teams using statistics. Sources reviewed 4 October 2026. Numerical examples are invented for education.
18+. Betting involves risk. No statistic or strategy guarantees profit. Only spend money you can afford to lose and never chase losses.
