Goals conceded can flatter or condemn a defence unfairly, because a team can concede few goals through luck or an inspired goalkeeper, or concede many despite defending well. Expected goals against, or xGA, solves much of that problem by measuring the quality of chances a team allows, rather than how many of those chances happened to be scored. It has become a key tool for judging defensive performance.

Every shot a team concedes is assigned an expected goals value based on distance, angle, defensive pressure and shot type, just as attacking xG values shots taken. Adding up those values over a match or a season gives xGA: the number of goals a team would typically concede given the chances it allowed. A team with low xGA is limiting opponents to few or poor chances, while a team with high xGA is allowing frequent or high-quality chances, regardless of the actual goals conceded.

The gap between goals conceded and xGA reveals luck, goalkeeping, or finishing quality. A team conceding far fewer goals than its xGA may have an excellent goalkeeper or fortunate opponents, and its defensive record may worsen as luck evens out. A team conceding more goals than its xGA may have suffered poor goalkeeping or exceptional opponent finishing, and its defence may be better than the table suggests.

Goals conceded remain the ultimate measure, because matches are decided by actual goals, not expected ones. But xGA is a better predictor of future defensive performance, because it strips out the randomness of finishing. A team that consistently allows low xGA is likely to keep conceding few goals over time, while a team allowing high xGA is living dangerously even if results have been kind so far.

Analysts therefore use both together: goals conceded for what happened, xGA for how sustainable it is. A title challenge built on low xGA is usually solid, while one built on low goals conceded despite high xGA may be fragile. The same logic applies to relegation battles, where a team with good xGA but poor results may be closer to safety than the table shows.

Low xGA comes from defensive organisation, not just individual defenders. Teams that press effectively prevent opponents from building attacks, teams that defend compactly limit space in dangerous areas, and teams that control possession deny opponents the ball altogether. Goalkeeper positioning and shot-stopping also affect xGA indirectly, because a goalkeeper who claims crosses and sweeps behind his defence prevents shots from occurring at all.

Set piece defending is a major component: teams that concede many chances from corners and free kicks will have high xGA even if their open-play defence is solid. Coaches therefore analyse xGA by phase, open play versus set pieces, and by zone, to identify whether problems come from transitions, wide areas, or dead balls.

When debating whether a defence is genuinely good, check xGA alongside goals conceded and clean sheets. If all three point the same way, the verdict is clear. If they diverge, ask why: is the goalkeeper overperforming, are opponents missing big chances, or is the team defending well but unlucky? Over a season, xGA usually converges with reality, making it a useful early warning system.

Expected goals against does not replace watching a defence, but it adds a layer of fairness that goals alone cannot provide. It judges defences by the chances they allow, not by the luck of whether those chances were scored.

Expected goals against measures the quality of chances a team concedes, offering a fairer and more predictive view of defensive strength than goals conceded alone. Use it alongside actual goals to judge whether a defensive record is sustainable or living on luck.