Archive
How to Read xG and xA Like a Pro Scout
Goals and assists don't tell the whole story. Learn how to read xG and xA like a scout, from per-90 totals to sample size and role context.
Think goals and assists tell the whole story? They don’t.
xG (expected goals) and xA (expected assists) measure chance quality, the kind of chances a player gets or creates, not whether the ball actually hit the net.
In this quick guide I’ll show you how to read xG and xA like a pro scout: when totals matter vs per-90, how big a sample should be, why role and position change the numbers, and why different providers can move the needle.
Read smart, not loud.
Quick Guide to xG and xA in Player Statistics

Higher xG means a player gets on the end of more shots, or better quality shots, than most. Higher xA means that same player is setting teammates up with genuinely good scoring opportunities, not just racking up pass counts. Neither stat is a prophecy. A player sitting on 0.45 xG per 90 isn’t guaranteed to score. It just tells you the chances falling to him have historically been converted at a high rate, and that’s a different thing entirely.
The totals versus per-90 split matters more than people think. A raw xG total shows accumulated output over a season, rewarding players who stay on the pitch and keep getting minutes. Per-90 numbers strip the playing time out of the equation, so you can compare a player with 900 minutes against one with 2,700 minutes on a level footing. Just make sure the sample is big enough first. Three good matches don’t make a per-90 rate trustworthy yet.
| Statistic | Plain-Language Meaning | Immediate Reference Range |
|---|---|---|
| xG | The combined scoring probability of a player’s shots, interpreted relative to position and minutes played | Varies by role and sample size |
| xA | The combined scoring probability of shots resulting from a player’s passes, interpreted relative to creative role | Varies by role and sample size |
| xG/90 | Expected scoring output per 90 minutes | Forwards: above 0.40 elite, 0.25-0.40 good, below 0.20 low. Wingers/attacking mids: roughly 0.10-0.25 |
| xA/90 | Expected chance creation per 90 minutes | Above 0.20 is a useful high-output reference |
Want to check these numbers yourself? Understat and FBref are free and easy to browse. StatsBomb, Opta, and Wyscout sit on the professional side, often behind club or media subscriptions. One quick piece of advice: stick to one provider when comparing players. Models disagree on the details, so mixing sources can make a perfectly good player look worse, or better, than he actually is.
Pros and Cons of xG and xA Player Analysis
(((alt-img2)))
xG and xA ask a better question than the scoresheet does. Instead of just logging whether a goal happened, they measure whether the chance itself was good. That distinction is the whole point, and it’s also where the limitations start creeping in.
| Advantages | Limitations |
|---|---|
| Measures chance quality, not just outcomes | Provider models differ, so values aren’t always directly comparable |
| Per-90 adjustment normalizes for playing time | Small samples produce volatile, unreliable numbers |
| Helps identify sustainable trends versus hot streaks | Tactical role, opponent quality, and league context all shift the numbers |
| Reveals value beyond raw goals and assists tallies | Off-ball movement and pressing contributions aren’t captured |
| Position-aware comparisons are more honest than raw totals | No single metric explains why one shot went in and another didn’t |
Here’s the part that trips people up most: regression to the mean. A striker running hot at 1.4 goals per xG for six matches is probably going to cool off, because most players eventually drift back toward their underlying rate. That’s not pessimism, it’s just how averages work over a larger sample. But sustained overperformance across a season or two, especially from proven finishers, can reflect real technical quality rather than random variance. The honest answer is you need both the stat and the patience to watch it over time. Treat xG as a strong signal, not a verdict carved in stone.
What Expected Goals Reveal About Player Performance
(((alt-img3)))
xG sits on a scale from 0 to 1, and the number tells you how often a shot like that one has historically found the net. A tap-in from two yards out might carry 0.90 xG. A speculative effort from 30 yards with three defenders in the way might sit closer to 0.05. A well-placed shot from the edge of the box could land around 0.30. None of these numbers judge the actual shot that was taken. They describe the history of similar attempts.
Providers build these models by feeding in hundreds of thousands of historical shots and tracking what separates goals from misses. “What is xG?” walks through how these historical patterns get translated into a single probability. The main variables that go into the calculation include:
- Distance from goal
- Shooting angle
- Foot, head, or other body part
- Open play, penalty, or set piece
- Through ball, cross, cut-back, or rebound
Final Words
When a player reaches the box, xG shows the quality of the shot and xA shows the quality of chances set up for teammates. They measure chance value, not guaranteed goals or assists.
Totals show accumulated output; per-90 lets you compare minutes fairly. Remember position benchmarks and beware tiny samples.
Use Understat or FBref, stick to one provider, and pair numbers with video — that’s the core of how to read xG and xA for player performance analysis. It’s a practical place to start.
FAQ
Q: What is xG and xA?
A: xG (expected goals) estimates the probability a shot becomes a goal, while xA (expected assists) sums the xG value of shots created by a player’s passes, both measuring chance quality, not guaranteed outcomes.
Q: How do totals differ from per-90 stats?
A: Totals show accumulated contribution across minutes; per-90 normalizes output to 90 minutes so players with different playing time become comparable, assuming you have a meaningful sample size.
Q: What are practical benchmarks for xG/90 and xA/90 by position?
A: Forwards: xG/90 >0.40 elite, 0.25–0.40 good, <0.20 low. Wingers/attacking mids: ~0.10–0.25 xG/90. xA/90 above 0.20 indicates strong creative output.
Q: How is xG calculated?
A: xG is trained on large shot histories and estimates scoring probability from variables like distance, angle, body part, shot type, rebound, defender pressure, goalkeeper position, and assist type.
Q: How is xA calculated and how does it differ from assists?
A: xA credits the passer with the xG value of the resulting shot—so a pass leading to a 0.35 xG attempt adds about 0.35 xA—separating chance quality from whether the teammate actually scored.
Q: What do normalized metrics like npxG, xG/90, and xGI mean?
A: npxG removes penalties; xG/90 and xA/90 adjust totals for minutes; xGI is xG plus xA (often per 90). They help compare players but can be inflated by small samples or role differences.
Q: What is post-shot xG and goals/xG ratio?
A: Post-shot xG evaluates shot placement after contact; goals/xG compares finishing to expectation. Ratios near 1.0 match expectation; >1.2 or <0.8 can flag over- or underperformance, given adequate sample size.
Q: What are xG and xA’s main limitations?
A: xG and xA measure chance quality but can mislead due to provider model differences, small samples, tactical role, league context, defensive pressure, and omitted off-ball contributions—use alongside video and context.
Q: Where can I find reliable xG and xA data?
A: Public sources like Understat and FBref provide accessible xG/xA. Professional or paid providers include StatsBomb, Opta, and Wyscout—keep comparisons within a single provider because models differ.
Q: How should scouts or fantasy managers use xG and xA?
A: Use xG/xA to spot under-credited scorers or creators, tracking per-90 trends, role-aware percentiles, minutes, and video. Treat rising xG/90 or xA/90 as an early signal, not an automatic decision.
Q: How do I compare two players using xG and xA?
A: Compare same league, season, and broad position; use npxG/90, xA/90, percentiles, goals-versus-xG lines, shot maps, and video to check tactical fit and whether differences stem from role or skill.



