What xG is trying to estimate
An expected-goals model assigns a probability to a shot using information such as location, angle, assist type and defensive pressure. Add those estimates together and we get a picture of the chances a team created.
Different models use different inputs, so two providers can publish slightly different totals for the same match.
Three interpretation traps
A higher xG total does not automatically mean a team deserved to win. Game state affects risk, low-probability shots can inflate volume and one exceptional chance may be more informative than several speculative efforts.
- Treating model output as an objective replay of the match
- Comparing numbers from different providers without context
- Using one match to make a long-term claim about finishing
Use the number to ask a better question
The best use of xG is as a prompt: where were the chances created, what happened before the shot and did the pattern repeat? The model points us toward the film; it does not replace it.
This page is part of the Casino Review Daily structural prototype. Market-specific facts must be verified against primary sources before publication.