Standard xG tells you the quality of the position a player shot from. Post shot xG tells you the quality of the shot they actually produced, and that distinction is what separates chance creation metrics from shot-stopping metrics.
By David Findlay, Founder of KiqIQ.
Quick Answer: Post shot xG (PSxG) measures the probability of an on-target shot resulting in a goal by combining the underlying chance quality with the actual location the ball travels to within the goalmouth. It applies only to shots on target and is the primary metric for evaluating goalkeeper shot-stopping performance.
Definition: PSxG is calculated after the ball leaves the player’s foot, incorporating shot trajectory and the estimated y and z coordinates of where the ball would enter the goalmouth. These variables serve as a direct measure of save difficulty, independent of the position the shot was struck from.
Key point: Standard xG covers all shot attempts, including off-target efforts. PSxG is calculated exclusively for shots on target: those that require a goalkeeper intervention. A shot hit straight at the keeper from a strong position will have a lower PSxG than the same shot placed toward the top corner.
How PSxG Differs From Standard xG
Standard xG is a pre-shot metric. It evaluates the probability of a goal based on information available at the moment the player strikes the ball: location, angle to goal, body part used, and delivery type. It does not know where the ball actually goes.
PSxG is a post-shot metric. It takes the underlying chance quality and layers in information about the ball’s actual trajectory and goalmouth destination. According to StatsBomb, the y and z location where the ball enters the goal mouth are among the most important factors for predicting goal probability, and provide a clear proxy for save difficulty.
A shot from a strong central position that is struck directly at the goalkeeper will register a high xG but a low PSxG: the position was good, but the execution reduced the difficulty for the keeper. A shot from the same position placed toward the top corner will produce a high PSxG, reflecting the actual difficulty of the save required.
4 Reasons PSxG Changes Goalkeeper Evaluation
- It separates defence from individual shot-stopping. Traditional save percentage metrics do not distinguish between a goalkeeper facing 10 difficult placements and one facing 10 efforts straight at them. PSxG assigns a difficulty value to each shot based on placement, isolating individual goalkeeper contribution from the quality of the defensive structure in front of them.
- It produces a goalkeeper performance benchmark. The difference between PSxG conceded and actual goals conceded, sometimes labelled Goals Saved Above Average (GSAA), gives an objective measure of shot-stopping performance over a period. A keeper allowing fewer goals than the PSxG total against them is saving shots they would not be expected to save based on placement alone.
- It neutralises shot volume distortion. A goalkeeper facing a high volume of low-difficulty shots may accumulate impressive raw save totals but a modest PSxG rating. PSxG-based evaluation accounts for this, rewarding keepers who stop genuinely difficult placements rather than those who face more speculative efforts.
- It removes goalkeeper positioning from the metric by design. The StatsBomb PSxG model does not incorporate goalkeeper positioning data. This is intentional: positioning quality is treated as a separate analytical domain. PSxG focuses specifically on shot-stopping once the ball is in flight.
PSxG for Outfield Evaluation
PSxG is not exclusive to goalkeeping analysis. Comparing a shooter’s PSxG total to their standard xG total reveals placement quality. A player whose shots consistently generate higher PSxG values than their pre-shot xG suggests their shot placement is adding probability above the positional advantage alone.
Over a season, this comparison distinguishes technical shot accuracy from positional shooting proficiency. A player in excellent positions who places shots centrally will accumulate high xG but lower PSxG. A player who consistently finds difficult corners will show a positive PSxG relative to xG.
Where to Access PSxG Data
FBref provides PSxG data at player and team level for hundreds of leagues, powered by Stats Perform (Opta). The metric appears in shot-level data tables and goalkeeper performance summaries, labelled PSxG for on-target shots faced or produced.
StatsBomb’s free open data on GitHub includes shot-level event data with PSxG values for select competitions, including precise goalmouth coordinate data used in their model. This dataset is the most detailed free resource available for PSxG-based goalkeeper analysis in Python or R.
Frequently Asked Questions
What is post shot xG?
Post shot xG (PSxG) is a metric that measures the probability of an on-target shot resulting in a goal, based on both the pre-shot chance quality and the actual location the ball travels to within the goalmouth after it is struck.
How does PSxG differ from xG?
Standard xG is calculated before the shot is struck, using location, angle, body part, and delivery type. PSxG is calculated after, incorporating shot trajectory and goalmouth placement. PSxG only applies to shots on target, while standard xG covers all attempts including off-target shots.
How is PSxG used to evaluate goalkeepers?
PSxG-against is compared to actual goals conceded. A goalkeeper allowing fewer goals than their PSxG-against total is saving shots at a higher rate than placement difficulty would predict, indicating above-average shot-stopping performance.
What variables does the StatsBomb PSxG model use?
The StatsBomb model uses the estimated y and z coordinates of where the ball would enter the goalmouth, along with shot trajectory and speed. Goalkeeper positioning is explicitly excluded from the PSxG calculation to isolate shot-stopping from positioning quality.
Can PSxG be used to evaluate outfield players?
Yes. Comparing a player’s total PSxG to their total xG reveals shot placement quality. Shots that consistently carry higher PSxG than xG indicate a player is finding difficult areas of the goal beyond what their positional advantage alone creates.
Where can I find PSxG data?
FBref provides PSxG data at player and team level for hundreds of leagues, powered by Opta. StatsBomb’s free open data on GitHub includes PSxG values for on-target shots in select competitions, including goalmouth coordinate data.
Sources
- StatsBomb (Hudl): A New Way to Measure Keepers’ Shot Stopping: Post-Shot Expected Goals
- FBref (Stats Perform / Opta): xG Explained, FBref
- StatsBomb Open Data: Free Football Data from StatsBomb, GitHub
- American Soccer Analysis: What are Expected Goals (xG)?
- Opta Analyst (Stats Perform): What is Expected Goals (xG)?