Most match stats pages now show two numbers side by side: xG and xGOT. They look alike and are often mixed up, but they answer different questions. For betting, each one is useful in different places.
This guide explains what xGOT adds, how to use it to judge goalkeepers and finishers, where to find both numbers for free, and how to turn xG into match probabilities with our free xG and correct score calculator.
If you're new to expected goals, start with our complete guide to xG. This article assumes you know the basics.
Note: StatPair doesn't provide xG or xGOT. Real figures in this guide come from the linked sources. Everything else is a clearly marked hypothetical example.
TL;DR
- xG rates the chance, measured before the shot. Every shot gets a value.
- xGOT rates the shot, measured after it is struck, using where it ends up in the goal. Only shots on target get a value.
- Goals prevented = xGOT faced minus goals conceded. The cleanest public measure of a goalkeeper's shot-stopping.
- xGOT minus xG shows how much a player's shot execution added to their chances. It is noisy over short runs.
- For predictions, use xG. Put each team's xG into the xG calculator to get 1X2, BTTS, over/under and correct score probabilities.
xG in one paragraph
xG gives each shot a probability of being scored, based on similar shots in the past: distance, angle, body part, type of assist and so on. Opta's model uses nearly one million historical shots (Opta Analyst stats definitions). It is measured at the moment the shot is taken, so it doesn't care whether the ball flew into the top corner or over the bar.
That is its strength and its limit. xG tells you how good the chance was. It tells you nothing about what the player did with it.
What is xGOT (expected goals on target)?
xGOT, also called post-shot xG (PSxG on some sites), is the probability that an on-target shot goes in, measured after the shot is struck. It combines the original xG with where the ball ends up in the goal mouth. A shot heading for the corner gets more credit than one hit straight at the goalkeeper (Opta Analyst: what are expected goals on target).
Three rules follow from that:
- Only shots on target get an xGOT value. A shot off target can't score. In Opta's model, blocked shots get zero too.
- xGOT can be higher or lower than xG for the same shot. Good placement raises it, a tame finish lowers it.
- A team's total xGOT can sit above or below its xG. Above means good shots from modest chances. Below means missed targets or weak finishes.
Same chance, three different shots (hypothetical)
| Shot | xG | xGOT | What it shows |
|---|
| Side-footed into the bottom corner | 0.30 | 0.65 | Excellent execution |
| Hit straight at the goalkeeper | 0.30 | 0.10 | Tame finish |
| Fired over the bar | 0.30 | None | Off target, so no xGOT |
All three players had the same chance, so xG rates them equally. xGOT separates them. The numbers are made up, but the pattern matches how Opta describes the model.
xG vs xGOT: the key differences
| xG | xGOT |
|---|
| Question it answers | How good was the chance? | How good was the shot? |
| Measured | Before the shot | After the shot |
| Which shots | All shots | Shots on target only |
| Main use | Chance creation and chances conceded | Shot execution and goalkeeper performance |
| Best betting use | Pre-match match odds, totals, correct score | Goalkeeper analysis, testing hot streaks |
An easy way to remember it: xG is about the chance, xGOT is about the shot.
Goalkeepers: goals prevented
Once a shot on target has been struck, only the goalkeeper stands between the ball and the net. That makes xGOT a fair test for keepers:
Goals prevented = xGOT faced − goals conceded (excluding own goals)
A real example: Opta Analyst reports that in the 2019/20 Premier League season Dean Henderson faced 39.4 xGOT for Sheffield United and conceded 32 goals (excluding penalties and own goals). That credits him with more than seven goals prevented, the fourth-highest total in the league that season (Opta Analyst).
Clean sheets and save percentage depend heavily on the defence in front of the keeper. Goals prevented strips most of that out. Two cautions:
- Use rates, not totals. A keeper who faces lots of shots has more chances to "prevent" goals. Compare per 90 minutes or per shot faced.
- Samples are small. A keeper might face only a handful of shots on target per match, so two or three saves swing the number. Treat ten games as a hint and a full season as evidence.
Finishers: xGOT minus xG
The same idea works for attackers:
Shot execution = xGOT − xG
Opta calls this "shooting goals added". If it's positive, a player's shots are ending up in better places than an average player would manage from the same chances. If it's negative, they miss the target or hit tame shots.
Opta adds a caveat: blocked shots get zero xGOT, so defenders affect the number. It measures on-target execution rather than pure finishing ability.
A third comparison is useful too: goals minus xGOT. If a striker scores well above their xGOT, keepers have been letting in shots they would usually save. That part is the least likely to last.
| Pattern | What it suggests | Betting angle |
|---|
| Goals > xGOT > xG | Good shots plus some goalkeeper errors | Expect cooling; goalscorer prices may be too short |
| Goals ≈ xGOT > xG | Strong shot placement | Some skill, but expect partial regression |
| Goals < xG, xGOT < xG | Poor execution or bad luck | Possible value if chance volume holds |
| Goals ≈ xGOT ≈ xG | Average finishing | Price on chance volume alone |
Hot finishing runs usually fade towards the player's long-term level. Our guide to regression to the mean explains why and how to spot it.
What this means for betting
1. Match odds, totals and correct score: use xG
Pre-match models should use xG for and against, because chance creation is more stable from week to week than finishing. xGOT includes shot execution, which swings a lot over short runs. Put each team's expected goals into the xG calculator to turn them into probabilities.
2. Goalkeeper saves
Saves are roughly shots on target faced minus goals conceded. The biggest driver is how many shots on target the opponent produces. Goals prevented then tells you whether the keeper turns those shots into saves at a better or worse rate than average.
A rough estimate of saves for an average keeper is shots on target faced × (1 − average xGOT per shot on target). As a hypothetical example, an opponent producing 5 shots on target with an average xGOT of 0.30 points to about 3.5 saves. Read our goalkeeper props guide for the full method.
3. Shots on target props
These markets are settled on the count, not the quality. xGOT tells you how good a player's shots on target were, not how many they will have. Shot volume, accuracy, minutes and the opponent matter far more. Our shots on target betting guide covers this. StatPair shows shots and shots on target for each player's recent matches, with home and away filters.
4. Anytime goalscorer
Use xG per 90 as the starting point. Then use xGOT minus xG to check whether a scoring streak is backed by good shots or by goalkeeper mistakes.
Where to find xG and xGOT for free
| Source | xG | xGOT / post-shot |
|---|
| Understat | Yes, big five leagues plus Russia | No |
| FotMob | Yes, many leagues | Yes, in match and player stats for many leagues |
| SofaScore | Yes, many leagues | Varies by league; check the match and goalkeeper pages |
| Opta Analyst | In articles and tables | Used in articles |
| FBref | Removed in January 2026 | PSxG removed too |
Coverage changes, so check the match page before you rely on a number. Stick to one provider inside any comparison, because each one uses a different model. Our comparison of the best free xG websites covers each source in more detail.
Worked example: turning xG into match probabilities
Here is a hypothetical match. The home side is expected to create 1.6 xG and the away side 1.1 xG. These are made-up inputs, not data for real teams. Enter them into the xG and correct score calculator and you get:
| Market | Probability | Fair odds |
|---|
| Home win | 49.0% | 2.04 |
| Draw | 24.9% | 4.02 |
| Away win | 26.2% | 3.82 |
| BTTS Yes | 53.2% | 1.88 |
| Over 2.5 goals | 50.6% | 1.97 |
| Under 2.5 goals | 49.4% | 2.03 |
| Home team over 1.5 | 47.5% | 2.10 |
The three results add up to 100.1% only because of rounding. The five most likely scores:
| Score | Probability | Fair odds |
|---|
| 1-1 | 11.8% | 8.45 |
| 1-0 | 10.8% | 9.30 |
| 2-1 | 9.5% | 10.57 |
| 2-0 | 8.6% | 11.62 |
| 0-1 | 7.4% | 13.53 |
Even the most likely score lands only about once in every eight or nine matches like this. Correct score is a high-variance market, which our correct score betting guide covers in depth.
The calculator uses the Poisson distribution. The chance of a team scoring exactly k goals when its expected goals are λ is:
P(k) = e^−λ × λ^k / k!
The chance of a scoreline is the home probability multiplied by the away probability. For 1-1: P(home scores 1 | 1.6) × P(away scores 1 | 1.1) = 0.323 × 0.366 ≈ 11.8%. Fair odds are 1 divided by the probability.
What if you used xGOT instead?
Now suppose the home side's recent xGOT is 2.0 per game because its striker is on a hot run. Enter 2.0 against 1.1 and the home win jumps to 58.2% (fair odds 1.72), with over 2.5 goals at 59.9%. That is a big shift built on the noisiest part of the data. Unless you have good reason to think the finishing will last, stick with xG.
Comparing with the bookmaker
Suppose a bookmaker offers 2.30 on the home win and your fair odds are 2.04. The price is above your fair value. It is only real value if your 1.6 and 1.1 are better estimates than the market's, and bookmakers use xG too.
No xG for the match? Estimate it
The calculator includes a helper that uses a standard method: expected goals = attack strength × opponent defence strength × league average. It is a method, not StatPair data.
Hypothetical example: the league average is 1.35 goals per team per match, the home side scores 1.8 per home game, and the away side concedes 1.5 per away game. That gives (1.8 ÷ 1.35) × (1.5 ÷ 1.35) × 1.35 = 2.00 expected goals for the home side. Use xG averages instead of goals if you have them.
Limits of the model
Poisson assumes the two teams score independently. In real matches, that slightly underrates low-scoring draws such as 0-0 and 1-1. Models such as Dixon-Coles adjust for this. The bigger risk is the inputs: poor expected goals estimates give poor probabilities, however neat the maths looks.
FAQ
Q: What is the difference between xG and xGOT?
A: xG measures the quality of a chance before the shot is taken, and every shot gets a value. xGOT measures the quality of the shot after it is struck, using where it ends up in the goal, and only shots on target get a value. xG is about the chance, xGOT is about the shot.
Q: Is xGOT the same as post-shot xG (PSxG)?
A: Yes, in practice. Opta calls it expected goals on target (xGOT), and other providers have used post-shot xG or PSxG for the same idea. Values differ slightly between providers because their models differ.
Q: How do you calculate goalkeeper goals prevented?
A: Goals prevented = xGOT faced minus goals conceded, excluding own goals. A positive number means the goalkeeper conceded fewer goals than an average keeper would have from the same shots. Compare per 90 minutes or per shot faced, and use at least a season of data.
Q: Should I use xG or xGOT to predict a match?
A: Use xG. It measures chances created and conceded, which are more stable from match to match. xGOT includes finishing and shot placement, which swing a lot over short runs. Put each team's xG into a Poisson calculator to get 1X2, BTTS, over/under and correct score probabilities.
Q: Is there a free xGOT calculator?
A: You don't need one, because data providers calculate xGOT shot by shot and sites such as FotMob show the totals. What you can work out yourself is goals prevented (xGOT faced minus goals conceded) and shot execution (xGOT minus xG). For match probabilities, use the free StatPair xG calculator with xG inputs.
Bottom line
xG and xGOT are not rivals. xG tells you how good the chances were, so it is the right input for predicting match results, goals and correct scores. xGOT tells you what happened after the shot, so it is the right tool for judging goalkeepers and testing whether a finishing streak is real.
Use xG to build your expectations, xGOT to check the stories behind the scorelines, and the free xG calculator to turn expected goals into prices you can compare with the bookmaker.