Post-Shot Expected Goals Explained
Post-Shot Expected Goals, usually shortened to Post-Shot xG or PSxG, is one of the most useful concepts in modern football analytics because it answers a question traditional Expected Goals cannot fully address: what happened after the player struck the ball?
Standard Expected Goals evaluates the quality of a chance before the shot is taken. It considers factors such as shot location, angle, body part, assist type, defensive pressure where available, and the game context included in the model. Post-Shot xG goes one step further. It evaluates the quality of the shot after contact, using information such as where the ball travelled within the goal frame, its speed or trajectory where available, and whether the attempt forced the goalkeeper into a difficult save.
That distinction matters. A striker can receive a high-quality chance and shoot straight at the goalkeeper. Another player can receive a low-quality chance and whip the ball into the top corner. Both moments may look similar in a traditional xG table if we only focus on the shot’s starting position, but they are very different once the ball leaves the boot.
Post-Shot xG helps separate chance quality from execution quality. It gives analysts a sharper view of finishing, shot placement, goalkeeper performance and, increasingly, predictive modelling.
What Is Post-Shot Expected Goals?
Post-Shot Expected Goals estimates the probability that a shot on target becomes a goal, based on the characteristics of the shot after it has been taken.
Traditional xG asks: given the situation before the shot, how likely was this chance to become a goal?
Post-Shot xG asks: given where the shot actually went, how likely was this shot to beat the goalkeeper?
That difference is crucial. A shot from 12 yards in a central position may have a strong xG value because the chance itself is dangerous. But if the player scuffs it softly into the middle of the goal, the post-shot value will be much lower. Conversely, a shot from a narrow angle may begin as a low-xG attempt, but if it is struck powerfully across goal into the far top corner, its Post-Shot xG can be much higher than the original chance value.
Most Post-Shot xG models are only applied to shots on target. If the shot misses the goal completely, it usually receives a Post-Shot xG value of zero because the goalkeeper did not need to save it and it could not directly become a goal. Some modelling approaches handle blocked shots or woodwork attempts differently, but the core idea remains the same: PSxG evaluates the goal probability of the shot after its final placement is known.
This makes Post-Shot xG especially valuable for analysing two areas that traditional xG often leaves blurred: finishing quality and goalkeeping quality.
Post-Shot xG vs Expected Goals
Expected Goals and Post-Shot Expected Goals are related, but they measure different parts of the shooting process.
Expected Goals is a pre-shot model. It evaluates the chance. Post-Shot xG is a post-contact model. It evaluates the execution.
Imagine three shots from the same location near the penalty spot:
The first is side-footed into the bottom corner.
The second is hit straight at the goalkeeper.
The third is dragged wide.
A traditional xG model may assign broadly similar values to all three because the shooting opportunity is similar. Post-Shot xG will treat them very differently. The shot into the bottom corner will receive a high PSxG value. The shot straight at the goalkeeper will receive a lower value. The shot wide will usually receive zero.
This is why Post-Shot xG should not be seen as a replacement for xG. It is an additional layer. xG tells us whether a team created good opportunities. Post-Shot xG tells us whether those opportunities were struck in a way that made them difficult to save.
The relationship between the two can reveal a great deal.
If a player consistently records more Post-Shot xG than xG, it may indicate strong shot placement or above-average finishing execution. If a player’s Post-Shot xG regularly falls below their xG, they may be getting into good positions but failing to test the goalkeeper effectively.
For teams, the comparison can show whether attacking underperformance is caused by poor chance creation or poor shot execution. A side generating strong xG but weak PSxG is probably arriving in dangerous zones but not hitting the target with enough quality. A side with low xG but high PSxG may be relying on difficult shots being finished unusually well.
Why Post-Shot xG Matters for Finishing Analysis
Finishing is often discussed in vague terms. A player is called clinical when they score regularly and wasteful when they miss. But raw goals can be misleading because they combine opportunity, execution and variance.
Post-Shot xG gives us a more precise way to examine finishing because it separates the chance from the shot outcome.
Consider a forward who takes ten shots worth 3.0 xG. If those shots generate 4.2 Post-Shot xG, the player has improved the value of their chances through shot placement and execution. They have not merely received good chances; they have made the goalkeeper’s task harder than the average player would from those situations.
Now consider another forward with the same 3.0 xG but only 1.5 Post-Shot xG. That player has turned good chances into relatively saveable efforts, missed the target too often, or failed to place shots away from the goalkeeper.
This is especially useful because goals alone are noisy. A striker can score from a weak shot if the goalkeeper makes an error. Another can strike an excellent shot that is saved brilliantly. Post-Shot xG does not remove all randomness, but it gets closer to the technical quality of the attempt than goals or traditional xG alone.
It also helps identify different types of finishers. Some players add value through power and placement from difficult angles. Others consistently hit the target but rarely find the corners. Some generate high xG through movement and positioning but produce average post-shot outcomes. Those distinctions matter when analysing recruitment, player development and tactical fit.
A poacher who repeatedly receives high-value central chances may not need to add much post-shot value to be effective. A wide forward who shoots from tougher angles may need exceptional placement to maintain strong output. Post-Shot xG helps place finishing performance in that tactical context.
Using Post-Shot xG to Evaluate Goalkeepers
Post-Shot Expected Goals is also one of the strongest public-facing tools for goalkeeper analysis.
Traditional save percentage treats all shots on target similarly. A goalkeeper who saves ten weak shots straight at them can look better than one who faces four high-quality finishes into the corners. Post-Shot xG adjusts for the quality of the shots faced.
The key comparison is usually goals conceded versus Post-Shot xG faced.
If a goalkeeper faces shots worth 10.0 PSxG and concedes 8 goals, they have prevented roughly 2 goals relative to the average expectation of those shots. If they face 10.0 PSxG and concede 13, they have allowed more goals than expected based on shot placement.
This type of analysis is often described as goals prevented or shot-stopping performance. It gives a much fairer assessment than clean sheets or save percentage because it accounts for how difficult the saves were.
For example, a goalkeeper behind a dominant defensive team may face few shots, but the shots they do face could be clear one-on-ones or close-range efforts. Another goalkeeper may face many speculative attempts from distance. Without Post-Shot xG, it is difficult to compare them fairly.
PSxG does not capture every goalkeeping action. It does not fully measure cross claiming, sweeping, distribution, communication or positioning before the shot. But for shot-stopping specifically, it is far more informative than basic goals conceded.
It is also valuable because it can reveal hidden defensive issues. A team may appear defensively solid if they allow few shots, but if the shots they do allow consistently carry high Post-Shot xG, the goalkeeper may be masking structural problems.
How Post-Shot xG Is Modelled
A Post-Shot xG model estimates the probability that an on-target shot becomes a goal based on post-shot information.
The most important variable is usually shot placement within the goal mouth. Shots close to the corners, particularly high and wide of the goalkeeper’s reach, tend to be more dangerous. Shots down the centre, especially at comfortable height, are usually easier to save.
Other possible variables include shot speed, height, trajectory, distance travelled, body part, goalkeeper position, defensive pressure and whether the shot was a header or struck with the foot. The exact inputs depend on the data available. Event data often provides shot location and outcome, while more advanced tracking or optical data can provide richer detail about ball movement and goalkeeper positioning.
A simplified model might look at a shot that is on target and ask:
Where did it enter or cross the goal frame?
How far was it from the centre of the goal?
Was it high, low or mid-height?
Was the original shot from close range or far away?
Was it a header, volley or regular footed shot?
Was it from open play, a counterattack or a set piece?
The model then estimates how often similar shots become goals.
The best Post-Shot xG models are not just maps of the goal frame. They recognise that the same final placement can have different difficulty depending on the shot origin. A low shot into the corner from six yards is different from the same placement from 30 yards because the goalkeeper has more time to react to the long-range attempt. Similarly, a close-range header placed centrally may still be dangerous because of reaction time.
This is why Post-Shot xG should be understood as a probability model, not a simple label. It is not saying a shot “should” or “should not” be saved. It is estimating how often shots with similar characteristics are scored.
Practical Examples of Post-Shot xG in Match Analysis
Post-Shot xG becomes especially useful when paired with the match narrative.
Suppose Team A wins 2-0 and records 1.1 xG. At first glance, the result may look like efficient finishing or mild overperformance. But if their Post-Shot xG is 2.4, the picture changes. They may not have created many clear chances, but the shots they put on target were placed extremely well. The finishing quality explains much of the scoreline.
Now imagine Team B loses 1-0 despite producing 2.0 xG. If their Post-Shot xG is only 0.6, they did not simply get unlucky. They created chances but failed to turn them into dangerous shots on target. Their problem was not chance creation; it was execution.
For goalkeeper analysis, imagine a keeper concedes three goals from shots worth 2.9 Post-Shot xG. That does not necessarily imply poor performance. The goals were broadly consistent with the quality of shots faced. But if another keeper concedes three from 0.8 PSxG, that raises a different question. Were the shots poorly dealt with? Was there a deflection? Did the goalkeeper’s positioning leave too much of the goal exposed?
Post-Shot xG helps analysts avoid lazy conclusions. Not every defeat with high xG is unlucky. Not every goalkeeper who concedes multiple goals has played badly. Not every striker who scores has finished well. The metric sharpens the diagnosis.
How Football Data Lab Applies Post-Shot xG
Football Data Lab uses Post-Shot xG as a second-stage evaluation layer within the shot modelling pipeline. The purpose is not simply to label shots as good or bad, but to separate three connected components: chance creation, shot execution and goalkeeper intervention.
The process begins with the pre-shot xG estimate. This gives the baseline probability of a shot becoming a goal before contact is made. That baseline is driven by the chance environment: shot location, angle, body part, attacking move type and other available contextual features.
Once the shot outcome is known, Football Data Lab applies a post-shot adjustment for attempts that reach the target. The post-shot layer evaluates the realised danger of the attempt by incorporating the shot’s placement relative to the goal frame and, where the data supports it, contextual features that affect reaction time and save difficulty.
A simplified example illustrates the pipeline.
A striker receives a cut-back near the penalty spot. The pre-shot xG estimate is 0.30, meaning that chances of this type are scored around three times in ten under the model’s assumptions. The player strikes the shot low and close to the goalkeeper. The Post-Shot xG may fall to 0.18 because the execution made the chance easier to save than the average penalty-spot opportunity.
In another case, a winger shoots from a narrow angle with a pre-shot xG of 0.08. The ball is driven across goal into the far side-netting area. The Post-Shot xG may rise to 0.25 because, although the original chance was difficult, the placement significantly increased the probability of scoring.
Football Data Lab then uses the gap between xG and PSxG as an execution signal. At player level, repeated positive gaps suggest that a shooter is adding value after the shot decision. Repeated negative gaps suggest that promising opportunities are being under-hit, misplaced or missed entirely. This is not judged on one match. The signal is assessed over a sample because finishing and shot-stopping metrics are volatile in small quantities.
For goalkeepers, Football Data Lab compares Post-Shot xG faced with actual goals conceded. The difference is treated as a shot-stopping component rather than a complete goalkeeper rating. This distinction matters. A goalkeeper can prevent goals through saves while still creating risk through poor distribution or weak command of the area. Conversely, a keeper may concede close-range finishes that no realistic model would expect to be saved.
In prediction modelling, Post-Shot xG is not used as a direct replacement for pre-shot xG. Pre-shot xG remains more stable for forecasting team attacking and defensive quality because it reflects chance creation, which is generally more repeatable than finishing streaks. However, Post-Shot xG is valuable as a diagnostic modifier. If a team’s goal output is far above its xG but closely aligned with its PSxG, the model can distinguish strong recent execution from pure finishing luck. If goals exceed both xG and PSxG, the overperformance is treated more cautiously.
The same logic applies defensively. A team conceding fewer goals than expected may be protected by excellent goalkeeping, poor opposition finishing, or variance. By comparing xG conceded, Post-Shot xG conceded and goals conceded, Football Data Lab can identify whether the defensive record is built on suppressing chances, forcing weak shot execution, or relying on the goalkeeper to outperform the shot quality faced.
This layered approach improves interpretation. The model does not blindly reward teams for scoring more goals than expected, nor does it automatically punish teams for conceding from excellent finishes. Post-Shot xG acts as a bridge between the pre-shot chance model and the final scoreline.
Post-Shot xG and Match Prediction
Post-Shot xG has a careful role in prediction. It is tempting to treat it as a superior version of xG because it is closer to goals, but that can be misleading.
For forecasting, the most valuable metrics are those that are repeatable and predictive. Pre-shot xG is often more useful for estimating future team strength because teams have greater control over chance creation than over the exact placement of every shot. Post-Shot xG contains more information about what happened in a match, but it can also contain more finishing variance.
That does not make PSxG irrelevant. It helps explain whether recent results were driven by sustainable processes or exceptional execution.
If a team has scored heavily from low xG and low PSxG, the finishing run may be fragile. If the same team has low xG but high PSxG, the question becomes more nuanced. Are they relying on outstanding shot placement from a small number of attackers? Is that player quality repeatable? Are they creating low-volume but high-control shooting situations?
For goalkeepers, Post-Shot xG can identify whether a defensive record has been inflated by strong shot-stopping. A club conceding few goals despite allowing high PSxG may be vulnerable if goalkeeper performance regresses or if the backup keeper plays.
In match prediction, PSxG is best treated as explanatory evidence rather than a raw input to be over-weighted. It can improve the interpretation of goals, but it should be balanced against shot volume, pre-shot xG, team style, player availability and venue effects.
Common Misunderstandings About Post-Shot xG
The first misunderstanding is that Post-Shot xG measures whether a player “should have scored”. It does not. It estimates the probability of a shot becoming a goal after placement is known. A shot with 0.40 PSxG is still saved more often than not.
The second misunderstanding is that PSxG proves finishing skill over a small sample. A player can outperform xG and PSxG over a few matches through variance. Reliable conclusions require larger samples and contextual analysis.
The third is that every high-PSxG goal is unsaveable. Goalkeepers can save excellent shots, and they can concede weaker ones. PSxG measures probability, not destiny.
The fourth is that PSxG replaces video analysis. It does not capture every technical detail: disguise, spin, goalkeeper sightline, deflections, body shape, balance and pressure may be missing depending on the data. The metric is powerful, but it is strongest when combined with tactical and technical review.
Limitations of Post-Shot Expected Goals
Post-Shot xG is a major improvement over basic shooting and goalkeeping metrics, but it has limitations.
It usually ignores shots off target by assigning them zero. That is logical for measuring whether a shot could directly become a goal, but it can understate the difference between a shot narrowly wide and one blasted into the stands. For finishing analysis, off-target frequency still needs to be considered alongside PSxG.
Data quality is another issue. A model with precise shot placement, speed and goalkeeper position will be stronger than one with only basic event data. Different providers may produce different PSxG values because their inputs and modelling assumptions differ.
Sample size is also critical. Goalkeepers and forwards can appear exceptional or poor over short runs because goals are rare events. A few outstanding finishes or unusual errors can distort the picture.
There is also the question of defensive pressure. A shot placed centrally may look poor, but if the player was stretching under heavy pressure, the execution might be understandable. Similarly, a goalkeeper’s positioning may make one side of the goal more available, affecting where the shooter aims. Without tracking data, some of these dynamics remain hidden.
Post-Shot xG is therefore best used as part of a wider analytical framework, not as a single verdict.
FAQ
What does Post-Shot xG mean?
Post-Shot xG measures the probability that a shot on target becomes a goal based on what happened after the shot was taken, especially where the ball was placed within the goal frame.
How is Post-Shot xG different from xG?
Expected Goals evaluates the quality of the chance before the shot. Post-Shot Expected Goals evaluates the quality of the shot after contact. xG measures opportunity; PSxG measures execution.
Can Post-Shot xG be used for goalkeeper analysis?
Yes. By comparing Post-Shot xG faced with goals conceded, analysts can estimate whether a goalkeeper has saved more or fewer goals than expected based on the quality of shots faced.
Does a higher Post-Shot xG always mean better finishing?
Not always. A higher PSxG indicates that shots were more dangerous after placement, but finishing assessment also needs sample size, shot selection, off-target attempts and tactical context.
Why do missed shots usually have zero Post-Shot xG?
Because a shot that misses the target cannot directly become a goal. Post-Shot xG is generally focused on the probability of on-target shots beating the goalkeeper.
Is Post-Shot xG better than Expected Goals?
It is not better; it answers a different question. xG is usually more useful for measuring chance creation and forecasting future performance. Post-Shot xG is better for analysing shot execution and goalkeeper shot-stopping.
Conclusion
Post-Shot Expected Goals adds an essential layer to football analytics by measuring what traditional xG cannot: the quality of the shot after it leaves the player’s boot. It helps distinguish good chances from good finishes, poor shooting from poor luck, and difficult saves from routine ones.
For attackers, PSxG reveals whether players are improving or reducing the value of their chances through execution. For goalkeepers, it provides a fairer view of shot-stopping than save percentage or goals conceded. For prediction modelling, it acts as a diagnostic tool that helps separate repeatable team process from finishing and goalkeeping variation.
Used carefully, Post-Shot xG does not replace tactical analysis or traditional xG. It makes both sharper.
See how these numbers translate to real fixtures on our live match predictions page, updated daily throughout the season.