Soccer Stats Explained: xG, Possession, and What Actually Matters

Baseball was quantified decades before soccer because baseball is a sequence of discrete events with clear owners. A pitch happens, a batter succeeds or fails, and responsibility is unambiguous. Soccer is ninety minutes of continuous, overlapping actions where the most valuable contribution in a goal might be a run made by someone who never touched the ball.

Add the scoring rate and the problem compounds. A typical match produces around two and a half goals. That is a tiny sample of decisive events, which means results swing wildly on luck in a way a 162-game baseball season smooths out. A team can play well and lose; a team can be dreadful and win 1-0 on a deflection. Any statistic worth using has to grapple with that, and the good ones do exactly one thing: they look at the process rather than the outcome, because process is far more repeatable than a two-goal sample.

~2.7

goals in a typical professional match

~1,000

passes across both teams in a match

~12

shots per team, most from poor positions

Expected goals — xG — is the one genuinely important statistic soccer has gained. Every shot is assigned a probability of becoming a goal based on the characteristics of comparable historical shots: distance, angle, body part, whether it followed a cross, how many defenders were between the ball and the goal. A tap-in from six yards might be rated 0.7, meaning shots from that situation score about 70 percent of the time. A hopeful strike from thirty yards might be 0.03.

Add up every shot in a match and you get a team total. A scoreline of 1-0 sitting on top of an xG of 0.6 to 2.4 tells you something the scoreline hides: the losing team created considerably more, and if those teams played ten times the result would usually go the other way.

The other essential caveat is sample size. Over one match, xG is noisy and a single big chance dominates the total. Over fifteen or twenty matches it becomes genuinely predictive — better at forecasting the goals a team will score than that same team’s own past goals. A side massively outscoring its xG across a season is usually not sustaining it, and a side underperforming is usually about to improve. That is the whole reason analysts talk about it constantly.

Possession percentage is the oldest stat on the graphic and among the least informative. It measures how long a team had the ball, not what they did with it. A team can hold 70 percent possession by passing between center-backs against a side that is perfectly happy to sit deep and let them.

Some of the most effective teams in history have deliberately conceded possession, absorbing pressure and attacking the space left behind. Others build their entire identity on keeping the ball. Neither approach is inherently better, which is precisely why the number alone predicts so little. What matters is where the possession happens and what it produces — territory in the final third, entries into the penalty area, shots created.

Possession is not a goal in itself. It is a tool, and a team that does not want the ball has not necessarily lost anything.

Beyond xG, a small set of measures reliably adds something. These are the ones worth recognizing when they appear.

Notice the pattern. Each of these describes something a team is repeatedly trying to do, rather than a single outcome that may or may not have come off. That is the mark of a statistic with predictive value, and it is a useful filter when you meet an unfamiliar one.

Several long-standing numbers survive on broadcasts mainly out of habit, and a few actively point the wrong way.

• Total shots — treats a thirty-yard punt and a six-yard tap-in as equal events, which is exactly the error xG was invented to fix.
• Corners — weakly related to scoring. A team winning many corners is often being repelled effectively.
• Tackles — a defender who reads the game well makes fewer tackles because he intercepts instead. High tackle counts can indicate constant defending.
• Distance covered — a losing team chasing the ball covers more ground. Running further is not the same as working better.
• Pass completion — inflated by safe sideways passing. A creative midfielder attempting hard passes will always look worse.

League tables have their own quiet statistic: goal difference. It is not a tiebreaker so much as a better description of a team than points are, particularly early in a season. Points are lumpy, since a one-goal win and a five-goal win pay identically, while goal difference accumulates evidence every week.

A team sitting fourth with a goal difference of plus two is fourth because of narrow wins that will not keep landing. A team sitting ninth at plus twelve is usually better than its position and tends to climb. Combine that with cumulative xG difference and you have a genuinely useful picture of a season by about matchweek ten, well before the table itself settles down.

The point of all this is not to replace watching with reading numbers. It is to stop the numbers from misleading you, and to give you a way of checking whether what you felt while watching actually happened.

1. 1

   Watch the match first

   Form your own impression before you look at anything. Stats are best used to test a view you already hold, not to manufacture one.

2. 2

   Check xG for both sides

   If it contradicts the scoreline, that is the interesting part. Ask why: was one team clinical, or did it defend its box well?

3. 3

   Look at the shot map, not the total

   Where shots came from tells you far more than how many there were. Clusters near the penalty spot mean a team is creating properly.

4. 4

   Zoom out to ten matches

   Single-match stats are noise. Trends over ten or more are where the signal lives, for teams and players alike.

One last thing worth keeping in view: none of this transfers to the pitch you actually play on. Nobody is tracking your progressive carries at the park on Tuesday. The analytics revolution has been genuinely good for understanding the professional game, and completely irrelevant to whether you enjoyed your own match — which remains, by a distance, the more important number.