If you've watched any serious football analysis in the last five years, you've definitely heard someone mention expected goals, or xG as it's commonly abbreviated. It's become the darling metric of modern football analytics, but there's still a lot of confusion about what it actually means and whether it's worth paying attention to. Let me cut through the noise.
Expected goals is essentially a measurement of shot quality. Each shot a team takes gets assigned a probability of becoming a goal based on historical data—where the shot came from, the angle, the distance from goal, and other contextual factors. If a team's shots add up to an xG of 1.5 for a match, that means the quality of their shots would historically convert to about 1.5 goals on average. It's not predicting what will happen; it's evaluating what probably should have happened based on what we know about shooting.
Here's where it gets interesting: xG is genuinely useful, but not for the reasons casual fans think it is.
People often treat xG like a report card. "We had 2.1 xG but only scored one goal—we were unlucky!" or conversely, "We scored three goals on 0.9 xG—we were clinical!" Both of these statements miss the actual point. Sure, randomness exists in football. A shot can hit the post, an opponent's shin, or sail in off the keeper's gloves. But if you're consistently creating chances with high xG and converting them at normal rates, you're doing something right. If you're consistently underperforming your xG, that's worth investigating, but it's usually either a problem with your finishing in training or a small sample size issue.
The real value of expected goals is that it reveals what kind of team you actually are, not what the scoreboard says you are.
This is where team quality becomes clear. A team that consistently generates high xG is building good attacking structure. Their movement off the ball is creating space. Their passing is finding dangerous areas. These are the things that matter in football. A team that regularly concedes low xG but gives up many goals has a defensive structure problem. Maybe their pressing is leaving gaps. Maybe their shape on the counter is wrong. Maybe they're not covering space efficiently. These problems will eventually show up in the results, even if they haven't yet.
Consider a manager looking at their team's numbers. If the data shows they're creating 1.8 xG per game but scoring 0.9, the easy excuse is "our strikers aren't clinical enough." But is that really true? Or is the quality of the chances actually reflective of a deeper issue? A really good striker will occasionally miss sitters, but they're unlikely to miss them consistently. If you're analyzing the data properly, you'll see whether the chances are actually "sitters" or just chances that the analytics model rates highly because of historical conversion rates for similar shots. Maybe your strikers aren't getting into truly high-percentage situations. Maybe your build-up play is settling for okay chances rather than great ones.
This is where game analysis becomes essential. The numbers tell you something is off, but you need to watch the tape to understand what. xG points you toward the problem; it doesn't solve it.
The other direction matters equally. If a team is consistently overperforming their xG—scoring more than their shot quality suggests they should—that's actually a warning sign, not a cause for celebration. It means luck is involved. It means either your shots are better quality than the model accounts for (which is possible but worth investigating) or you're genuinely fortunate with bounces and finishes. This is unsustainable. Eventually, regression to the mean happens, and those overperforming teams suddenly find the goals drying up. This is one of the most reliable predictors of a team's second-half performance. A team that's overperforming xG by a significant margin in the first half usually underperforms it in the second half, sometimes dramatically.
What's particularly useful about xG when evaluating team quality is how it helps you separate narrative from reality. We all want stories in football—a scrappy underdog overcoming odds, a fall from grace, a surprising emergence. But those narratives can cloud judgment. The numbers keep you honest.
Take the scenario of a mid-table team that plays "defensively" and catches teams on the counter. They might score impressive goals and hold a mid-table position while playing only occasionally. The narrative is "they're pragmatic and well-organized." But look at their xG for and against. If they're generating very low xG on both ends, what they actually are is lucky. They're benefiting from fine margins. The moment those margins shift, they'll collapse. If they're conceding reasonable xG but their keeper is having an exceptional season, that's sustainability—they have a good defensive structure backed by good goalkeeping. These are different situations entirely, and xG helps you distinguish between them.
The same applies in reverse. A team playing beautiful football and dominating possession might generate high xG but lose matches. The narrative might be "they're unlucky" or "the league is against them." But here's the thing: if they're consistently generating high-quality chances and not converting them, and it's not because their strikers are underperforming at the individual level, then maybe the problem is they're creating these chances at such a late stage of the attacking move that the opposition has time to recover. Maybe they're not being clinical enough in their positioning. The high xG indicates they're on the right track structurally, but something in the execution is off.
One more thing that's crucial to understand: xG works best when you're comparing teams over a season, not a single match. A team might have a fluke result where they score four goals on 0.8 xG. That's noise. When you zoom out and look at twenty or thirty matches, the true quality of the team's attacking and defending emerges.
The best coaches and analysts use xG as one data point among many. It's not the whole story. It doesn't account for pressing intensity, shape, set-piece organization, or countless other things that matter in football. But it's a genuinely useful lens for cutting through nonsense and understanding whether a team's structure is sound. It answers a fundamental question: "Are we actually a good team that's having a bad run, or are we a bad team that's having a good run?"
When you know the answer to that, you know what you're actually dealing with. That's the real value of expected goals.
Top comments (0)