Reading a Soccer Expected Goals Number Honestly

Expected goals is the most quoted number in soccer analysis and one of the most misread. It describes the quality of chances, not what should have happened, and that difference matters for every goal and shot prop.

By SlatelinePublished
A spread of shot locations, each marked with a small probability of scoring

Expected goals has become the default shorthand for whether a team or a player deserved more. A striker with a high xG and few goals is called unlucky, and a team that won with a low xG is said to have stolen a result. Some of that is fair. Much of it reads more into the number than the number says. For player props on goals, shots, and shots on target, reading xG honestly is worth more than reading it often.

This guide covers what xG measures, why different sources disagree, and how to use it in research without turning it into a forecast. It pairs with soccer player prop research, which covers the wider workflow.

What expected goals measures

The data provider Opta defines it this way in its explainer on expected goals: xG measures the quality of a chance by estimating the likelihood it will be scored, based on similar shots in the past. Each shot gets a probability between zero and one. A tap in from a few yards might sit near the top of that range, a speculative shot from distance near the bottom. Add the values across a match or a season and you get an expected goal total.

The inputs are features of the shot. Opta lists distance and angle to goal, the goalkeeper's position, and the type of shot, such as which foot was used or whether it was a header or a volley, and says its model weighs more than 20 variables in total. Some situations get fixed values. Opta assigns every penalty a constant 0.79 xG, reflecting the historical conversion rate of penalties.

The same explainer is careful about what xG is not. It measures chance quality, not the expected outcome of a match, and a team with more xG did not necessarily deserve to win. That one sentence is the key to reading the number well.

Why two sources give different numbers

There is no single expected goals. The Wikipedia entry on expected goals notes that xG is not one standardized statistic: providers train their own models on their own data with their own inputs, and the same shot can receive different values from different sources. Some models add defensive pressure from tracking data or the sequence of play before the shot, and others do not.

For research, that means two habits. Use one provider consistently when comparing players or matches, so you compare like with like. And treat a small difference between two players' xG figures as inside the margin of model choice, not as a real gap. The entry also carries a caution of its own: it is flagged for leaning on primary sources, so treat it as a starting point rather than an authority.

Small samples and the overperformance trap

Goals minus xG is the number behind most claims about finishing skill and luck. Over a single match it is almost entirely noise. Over a handful of matches it is still mostly noise. The Wikipedia entry notes that differences between goals and xG over small samples may reflect randomness, and that over and underperformance can reflect model choices as much as player skill.

That cuts both ways for props. A player scoring well above his xG over a short stretch is not proven to be an elite finisher, and a player well below it is not owed goals. The scoring run is a story. The underlying chances are the evidence. Streaks and slumps covers the general version of this problem, and sample size in player prop analysis explains how quickly short samples swing.

Example: Two strikers, same goals

For illustration, take two invented strikers who each scored five goals over ten matches. One produced about 4.5 xG from frequent chances close to goal. The other produced about 2.0 xG, mostly from long range shots that happened to go in. On a goals line, the first has the steadier foundation: his chances are good and frequent. The second's record depends on a conversion rate that the chances do not support. Neither is a certainty, and ten matches is not much evidence either way.

Using xG for goal and shot props

For goal props, per 90 minute xG is a reasonable starting point for how often a player gets good chances, provided you adjust for expected minutes. A player who often comes off around the hour mark gets fewer chances to add to his total, however good his rate. Soccer rotation risk covers how to estimate minutes when a manager rotates.

  • Shots and shots on target props depend on volume more than on chance quality. A player who shoots often from distance can carry a low xG per shot and still clear a shots line comfortably.
  • Penalties distort the picture. A player who takes penalties carries a large fixed xG on each one, so separate penalty xG from open play xG before comparing him with a teammate. Soccer set piece duties explains who takes them and why it matters.
  • Team context matters. A player's chances depend on how many his team creates, which depends on the opponent and the expected score.
  • Opponent quality shapes chances on both sides. A strong defense allows fewer good chances, which lowers everyone's expected output, not just the striker's.

What xG does not tell you

Expected goals does not know about the chances that never became shots. A move that ends in a misplaced final pass or a defender's last ditch block has no xG at all. It also does not tell you who will start, how a manager will set up, or whether a player is carrying a knock. Those questions often decide a prop more than the player's xG history does.

The honest way to use xG is as one input describing the quality and frequency of a player's chances, combined with minutes, role, set piece duties, and the opponent. Used that way it is a valuable piece of evidence. Used as a verdict on who deserved what, it becomes a story with a number attached.

References

Follow Soccer on Slateline

The public Soccer page lists each market with its graded record where one is published and, when a slate is open, the props the free plan opens. No account is needed to look.

Open the Soccer page

Not ready to sign up? Get the weekly note.

The Slate

One weekly note: what published, what the graded record shows, what changed. No picks, and nothing is sent until you confirm your address.

Keep researching

sport guides

How Minutes and Lineups Shape Soccer Player Props

Every soccer stat you can buy a line on flows through one narrow gate: time on the pitch. Before shots, passes, or tackles mean anything, research has to answer whether the player starts, and then whether they are still out there after the hour mark.

sport guides

Set Piece Duties and Who Actually Shoots

Dead ball duty is one of the few genuinely concentrated roles in soccer: two or three players inherit a stream of chances that has nothing to do with open play. It moves several prop families at once, and it changes without announcement.

sport guides

Rotation Risk in Congested Soccer Weeks

In most sports the biggest research risk is getting a rate wrong. In soccer it is getting the team sheet wrong. A player who is rested rather than injured is perfectly fit, fully available, and simply not playing, and no amount of accuracy about his shot volume survives that. Here is how to think about rotation before the lineup lands.

methodology

Streaks and Slumps: What a Hot Stretch Is Evidence Of

Form is the most persuasive story in prop research and the least reliable. The useful question is never whether a player is hot, it is whether something actually changed, because the streak is priced long before the mechanism is.

methodology

Why Sample Size Matters in Player Prop Analysis

Flip a fair coin ten times and seven heads is unremarkable. Watch a player for ten games and seven overs feels like destiny. The mathematics is identical; only the storytelling changes. This article covers how much evidence a sample really carries, why some stats settle down quickly while others take a season, and what to do when the data you have is all the data there is.

Reading a Soccer Expected Goals Number Honestly · Slateline