Non-Penalty xG Explained
Expected Goals has become one of the most useful concepts in football analysis because it separates finishing outcomes from the quality of chances created. A team can win 3-0 while producing very little, or lose despite creating the better openings. xG helps analysts look beyond the scoreline.
But not all xG is equally informative.
Penalties are the clearest example. A penalty is a high-value shot from a fixed location, with no defensive pressure in the usual sense, and with a conversion probability far above most open-play chances. If a player wins and scores a penalty, his xG total rises sharply. If a team receives two penalties in one match, its attacking numbers can look dominant even if it created very little from normal attacking play.
That is why analysts often use non-penalty expected goals, usually written as non-penalty xG or NP xG. It is not a replacement for xG. It is a more specific version of it, designed to answer a different question: how much chance quality did a team or player generate without penalties?
For player evaluation, team strength analysis, tactical assessment and prediction modelling, that distinction matters.
What Is Non-Penalty xG?
Non-penalty xG is simply expected goals with penalties removed.
If a team produces 2.1 xG in a match, but 0.8 of that came from a penalty, its non-penalty xG would be 1.3. The penalty still mattered in the match. It still changed the scoreline, the game state and the tactical environment. But for evaluating repeatable attacking quality, analysts often separate it from the rest of the team’s chance creation.
The reason is straightforward: penalties are not created in the same way as most shots.
A cut-back from the byline, a through ball behind the defence, a header from a corner and a long-range shot all tell us something about a team’s attacking structure. They reveal how a side moves the ball, accesses dangerous areas, disrupts defensive shape and creates shooting opportunities. A penalty tells us something different. It may result from pressure in the box, a dribble, a defensive error or a handball. But the shot itself is standardised in a way normal chances are not.
That standardisation is exactly why penalties can distort xG totals.
A player who takes penalties for his club may accumulate much more xG than a similar player who does not. A team that receives penalties at an unusually high rate over a short sample may appear more dangerous than its open-play chance creation suggests. Non-penalty xG helps strip out that effect.
Why Analysts Separate Penalties From Expected Goals
Penalties are valuable, but they are not always predictive in the same way as other shots.
This is the key analytical point. The purpose of xG is not only to describe what happened. It is often used to estimate underlying performance. Analysts want to know whether a team is creating sustainable danger, whether a striker is consistently getting into good scoring positions, or whether recent results are likely to continue.
Penalties complicate that analysis because they are relatively rare and unevenly distributed. A team might receive three penalties in five matches, then none for the next ten. A forward might score five penalties in a season and appear to have elite shot volume, even though his open-play involvement is less impressive.
Separating penalties allows analysts to ask cleaner questions:
How often does a team create high-quality chances from open play and set pieces?
How much of a striker’s goal threat comes from repeatable movement rather than penalty-taking duties?
Is a team’s attacking output driven by sustained pressure or by a small number of high-value events?
Are recent xG numbers inflated by penalties that may not continue at the same rate?
Non-penalty xG is therefore especially useful when the objective is to measure process rather than outcome.
How Penalties Can Distort Player Evaluation
Player analysis is where non-penalty xG becomes particularly important.
Imagine two forwards. Player A has 10 xG across a run of matches, including four penalties. Player B has 8 xG with no penalties. On total xG, Player A appears to have generated more scoring expectation. But once penalties are removed, Player B may have the stronger non-penalty profile.
That matters because non-penalty xG is usually more connected to movement, positioning and attacking involvement. A forward who consistently receives passes in central areas, attacks the six-yard box and gets shots from dangerous locations is showing repeatable goal threat. A player whose xG is heavily penalty-driven may still be valuable, especially if he is a reliable penalty taker, but his open-play scoring prospects should be evaluated separately.
This distinction also helps when comparing players across teams. Penalty-taking is often a role assigned by the coach. It can depend on hierarchy, seniority, set-piece responsibility or specialist skill. Two players with similar open-play threat can have very different total xG numbers simply because one takes penalties and the other does not.
For recruitment, this is a major issue. A club interested in signing a forward should not treat penalty xG and non-penalty xG as equivalent evidence of attacking quality. Penalty skill has value, but it should not be confused with the ability to generate shots from open play.
How Non-Penalty xG Improves Team Analysis
At team level, non-penalty xG is useful because it gives a clearer picture of attacking structure.
A side that consistently produces strong non-penalty xG is usually doing something repeatable. It may be progressing the ball into dangerous zones, creating overloads, winning the ball high up the pitch, generating cut-backs, or sustaining pressure through set pieces. Those patterns are more informative than a short burst of penalty events.
Consider a match in which a team records 2.4 xG. At first glance, that looks like a dominant attacking display. But if 1.5 of that total came from two penalties, the non-penalty xG might be only 0.9. That tells a different story. The team may still have deserved credit for forcing defensive errors in the box, but its chance creation from normal attacking sequences was modest.
The opposite can also happen. A team might lose 1-0 while producing 1.8 non-penalty xG from multiple high-quality chances. That performance is far more encouraging than the result suggests. Non-penalty xG helps analysts avoid overreacting to final scores and focus instead on chance quality generated through normal play.
It is also useful defensively. A team that concedes many penalties can have poor total xG against, but analysts need to distinguish between two problems: allowing dangerous shots from open play and repeatedly committing fouls or handballs in the box. Both matter, but they may require different tactical or coaching solutions.
Non-Penalty xG and Repeatability
The deeper reason non-penalty xG is valuable is repeatability.
Football analysis often tries to separate signal from noise. Goals are noisy because finishing, goalkeeping, deflections and small sample sizes can swing outcomes dramatically. xG improves on goals by measuring chance quality. Non-penalty xG goes one step further by reducing the influence of a high-value event that can arrive irregularly.
This does not mean penalties are random. Teams that attack the box frequently, use aggressive dribblers and force defenders into emergency actions may earn more penalties over time. But penalty counts are still volatile. A side can spend weeks getting into the box without receiving one, then win two in a single match.
Non-penalty xG is often better suited to assessing the stable parts of attacking performance: shot volume, shot location, assist type, centrality, defensive pressure, transition opportunities and set-piece threat. These elements can be connected to tactical behaviour and player roles.
For example, if a team’s non-penalty xG rises after switching from a slow possession system to a more direct pressing approach, that suggests the tactical change is creating better chances. If total xG rises only because of penalties, the interpretation is more cautious.
Where Non-Penalty xG Can Be Misleading
Non-penalty xG is powerful, but it should not be used carelessly.
The most common mistake is treating penalties as irrelevant. They are not. Penalties win matches. Drawing fouls in the box is a skill, and some teams are designed to create the kind of situations where defenders make mistakes. Wingers who isolate full-backs, forwards who receive with their back to goal in the penalty area, and midfield runners who arrive late into crowded boxes can all increase the likelihood of penalty incidents.
Another mistake is assuming non-penalty xG equals open-play xG. It does not necessarily. Non-penalty xG usually includes all shots except penalties, which means it can include corners, free kicks, throw-in attacks and other set-piece situations. If the analytical question is specifically about open-play creativity, analysts may need a separate open-play xG measure.
Context also matters. A team leading 2-0 may stop attacking aggressively, reducing its non-penalty xG. A team chasing the game may generate more shots because the opponent is defending deep. Red cards, fixture congestion, opponent quality and game state all influence chance creation.
Non-penalty xG is not a magic number. It is a cleaner measurement, not a complete explanation.
How Non-Penalty xG Feeds Into Match Predictions
Expected goals are central to many football prediction models because they provide a bridge between past performance and future scoring expectations.
A basic model might estimate how many goals each team is likely to score by looking at attacking and defensive xG rates. More advanced models adjust those rates for opponent strength, home advantage, match state, squad availability, rest, tactical style and competition context. Once expected goals for each team are estimated, they can be converted into scoreline probabilities.
Non-penalty xG is especially useful in this process because it helps estimate the repeatable scoring component. If a team’s recent attacking numbers are inflated by penalties, using total xG without adjustment may overstate its future goal expectation. Conversely, a team with strong non-penalty xG but poor finishing may be undervalued if the model relies too heavily on goals.
A prediction model should still account for penalties. The better approach is usually not to discard them completely, but to model them separately. Non-penalty chance creation and penalty expectation are different processes. A team’s expected scoring output can be thought of as the sum of its non-penalty goal expectation plus its penalty-related expectation.
That distinction allows a model to avoid two errors: ignoring penalties entirely, and treating them as if they are the same as open-play chances.
How Football Data Lab Applies Non-Penalty xG
In Football Data Lab’s modelling pipeline, non-penalty xG is not used as a cosmetic variation of xG. It is used to separate the scoring process into components that behave differently.
The first layer estimates each team’s non-penalty attacking and defensive strength. This is based on the quality and quantity of chances created and conceded excluding penalties, with adjustments for opponent strength and venue. A team that produces 1.6 non-penalty xG against a strong defensive opponent is treated differently from a team that produces the same number against a weaker side. Likewise, conceding low-quality shots from distance is not evaluated the same way as allowing repeated central chances.
The second layer handles penalties as a separate event process. Rather than allowing one or two penalties to dominate a team’s attacking rating, the model treats penalty expectation independently from non-penalty chance creation. This matters because penalties are high-value but low-frequency. They can heavily affect a single match total while providing a weaker signal about future open-play chance creation.
A simplified example shows why this matters. Suppose Team A records 2.2 xG in a match, with 1.5 coming from two penalties and 0.7 from non-penalty shots. Team B records 1.5 xG, all from non-penalty chances. A surface-level reading says Team A had the better attacking performance. In the Football Data Lab framework, Team A receives credit for creating or benefiting from penalty situations, but its underlying non-penalty attacking rating is not upgraded as strongly as Team B’s. Team B showed more repeatable chance creation through normal attacking play.
This separation then feeds into match prediction. The model projects a team’s expected goals by combining its adjusted non-penalty attacking strength with the opponent’s adjusted non-penalty defensive strength. Venue effects alter those expected goals because home and away conditions influence shot volume, territory, refereeing environment, tactical aggression and defensive behaviour. Penalty expectation is then layered separately rather than being baked blindly into open-play output.
The final expected goal estimate is therefore not just “recent xG averaged forward”. It is a structured estimate of how a team is likely to create chances in the next match, how the opponent is likely to suppress them, and how much penalty-related scoring expectation should reasonably be included.
This is particularly important after unusual matches. If a team’s last five-game xG average is boosted by several penalties, Football Data Lab’s process prevents that spike from being overinterpreted as a broad attacking improvement. If a team is consistently producing strong non-penalty xG but not scoring, the model can recognise that the underlying chance creation remains healthy even if recent results look poor.
Non-Penalty xG for Tactical Analysis
From a tactical perspective, non-penalty xG helps identify whether a team’s attacking plan is actually working.
A possession-heavy side may dominate the ball but produce low non-penalty xG if it cannot enter the box or create shots from central areas. Its passing numbers may look impressive, but the chance quality tells a different story. Analysts might then look at whether the team lacks runners beyond the defence, whether its wide players are crossing from poor positions, or whether the midfield is too slow to break lines.
A pressing team might have less possession but strong non-penalty xG because it creates chances immediately after turnovers. In that case, shot quality reveals the value of its defensive aggression. The team’s best attacking moments may come not from long passing sequences but from winning the ball near goal against an unsettled defence.
For set-piece specialists, non-penalty xG can also capture a major source of repeatable threat. Corners and free kicks are not penalties, but they can be trained, structured and exploited. A team with strong non-penalty xG from set pieces may deserve credit for coaching and delivery quality, even if its open-play creation is ordinary.
The important step is to break the number down further. Non-penalty xG is a starting point. The next questions are where the chances came from, which players were involved, what phase of play created them, and whether the pattern is likely to continue.
Non-Penalty xG for Betting and Forecasting
For forecasting, non-penalty xG is valuable because markets and public opinion can overreact to goals and headline xG totals.
A team that wins with two penalties may be praised for attacking dominance, while a team that creates multiple open-play chances and fails to score may be labelled wasteful or unlucky. Non-penalty xG helps correct those narratives.
In betting and probability modelling, the danger is overfitting to recent results. If a side’s goal difference improves because of a short penalty run, a model that does not separate penalties may inflate future scoring projections. That can lead to overestimating win probability, goal totals or player scoring chances.
Good forecasting treats non-penalty xG as one of the strongest indicators of sustainable attacking and defensive performance, while still respecting the value of penalties as match-changing events. The balance is important. Penalties should be included in expected goals for a match forecast, but their short-term distribution should not be allowed to masquerade as a permanent change in attacking strength.
FAQ: Non-Penalty xG
What does non-penalty xG mean?
Non-penalty xG means expected goals excluding penalties. It measures the quality of chances created from all non-penalty shots, which may include open play and set pieces depending on the data definition.
Why remove penalties from xG?
Analysts remove penalties because they are high-value, relatively rare events that can distort player and team xG totals. Excluding them gives a clearer view of repeatable chance creation.
Is non-penalty xG better than normal xG?
It depends on the question. Total xG is better for describing all scoring chances in a match. Non-penalty xG is often better for evaluating underlying attacking performance, player goal threat and future scoring potential.
Do penalties still matter in analytics?
Yes. Penalties are extremely important. The point is not to ignore them, but to analyse them separately from other chances because they arise and behave differently.
Is non-penalty xG the same as open-play xG?
No. Non-penalty xG removes penalties only. Open-play xG usually removes penalties and set pieces. The exact distinction depends on the data provider or model.
How is non-penalty xG used in predictions?
Non-penalty xG helps estimate repeatable attacking and defensive strength. In stronger prediction models, penalty expectation is then added separately so that forecasts include penalties without allowing them to distort open-play ratings.
Conclusion
Non-penalty xG is one of the most practical refinements of Expected Goals. By removing penalties, analysts get a cleaner view of how teams and players create chances through more repeatable attacking actions.
It should not replace total xG in every context, and it should never be interpreted without tactical and match context. But for evaluating player goal threat, team attacking strength and future performance, non-penalty xG often provides the sharper signal.
The best analysis does not pretend penalties do not exist. It separates them, models them properly, and then asks the more important question: what does the team create when the penalty spot is taken out of the equation?
See how these numbers translate to real fixtures on our live match predictions page, updated daily throughout the season.