In the last decade, Expected Goals (xG) has revolutionised the way football analysts, managers, and fans evaluate attacking performance. Developed to quantify the quality of scoring opportunities, xG assigns a numerical probability to each shot, indicating the likelihood it would result in a goal. A tap-in from a few yards might carry a high xG of 0.8, while a 30-yard speculative strike might sit at 0.05.
Yet, while xG has become a cornerstone of modern football analysis, it has its fair share of critics. From coaches wary of overreliance on statistics to fans questioning its real-world application, xG is far from universally accepted. Here, we explore the main criticisms surrounding this widely cited metric.
Key Criticisms of xG
The main criticisms of xG often focus on oversimplification, small sample variance, and context limitations. Here’s a breakdown:
| Criticism | Explanation | Counterargument |
|---|---|---|
| Oversimplifies the game | Reduces football to numbers, ignoring tactical intelligence, off-ball movement, and player creativity | Useful for trend analysis over the long term; complements rather than replaces qualitative assessment |
| Different xG models yield different results | Opta, StatsBomb, Wyscout, etc., all use slightly different algorithms | While models vary, overall trends across a season are reliable; differences are less important at scale |
| Ignores context and game state | High or low xG in desperate moments may misrepresent team performance | Contextual metrics like xGA (expected goals against) and shot quality over time can add nuance |
| Misrepresents finishing ability | Does not fully account for exceptional strikers or poor finishers | Best used for team-level evaluation; individual finishing can be evaluated alongside xG |
| High variance in small sample sizes | Single games can produce misleading xG stats | Most effective when analyzed over full seasons or multiple games |
| Ignores defensive quality | Focuses mainly on attacking, underplaying defensive tactics | xGA and other defensive metrics can complement xG to give a fuller picture |
1. xG Can Oversimplify the Game
One of the most frequent criticisms is that xG reduces the complexity of football to numbers. Critics argue that football is not just about chances—it involves tactical intelligence, off-ball movement, player creativity, and game context, which a simple xG value cannot capture.
For instance, a goal scored from a seemingly low xG chance might be the result of a brilliant individual effort or a perfectly timed run. Conversely, a high xG opportunity that is missed might reflect not poor finishing, but the pressure of a critical match moment. Analysts like Michael Cox have argued that overreliance on xG risks valuing opportunity quantity over quality of play, potentially ignoring subtleties that make football compelling.
2. Different xG Models Yield Different Results
Another major critique is that xG is not standardized. Multiple analytics companies—Opta, StatsBomb, Wyscout, and others—use slightly different algorithms and weightings. Some consider player position, defensive pressure, and shot type; others rely primarily on distance and angle.
This means the same match could have different xG totals depending on which model you use, raising questions about consistency and reliability. Critics suggest that while xG can indicate trends, it should not be treated as an exact measure of performance, especially in cross-comparisons between leagues or teams using different data providers.
3. Context and Game State Are Often Ignored
xG typically evaluates shots in isolation, without accounting for the context of a game. For example, a team trailing 3–0 late in a match might take low-quality, desperate shots—each with a tiny xG—but these don’t reflect the team’s overall attacking ability during the game.
Similarly, tactical context matters. A high-possession team dominating weaker opposition might generate large xG totals, but that doesn’t automatically equate to superior skill or strategy. Critics argue that without accounting for match state, pressure, and game circumstances, xG can provide a skewed picture of performance.
4. It Can Misrepresent Finishing Ability
One of xG’s most cited limitations is that it does not fully account for a player’s finishing skill. Legendary strikers often outperform their expected goals consistently, while others underperform. This discrepancy can lead to misjudgments:
- Overvaluing underperforming strikers: A forward who consistently misses high xG chances might be seen as unlucky, even if their finishing is poor.
- Undervaluing elite finishers: Players like Erling Haaland or Mohamed Salah often convert low xG chances at a rate far above statistical expectation. Relying solely on xG could downplay their clinical ability.
Critics suggest that while xG is great for evaluating team-level attacking performance, it should not replace qualitative assessments of individual skill.
5. High-Variance in Small Sample Sizes
xG is most reliable over long-term datasets, such as an entire season. In small sample sizes—single matches or short tournaments—variance can be extreme. A team might generate a high xG but fail to score due to an inspired goalkeeper, poor finishing, or just bad luck.
This is particularly problematic for media narratives, where single-game xG stats are often misinterpreted. Fans reading that their team “should have scored five goals” may feel aggrieved, even if the game outcome was realistic. Critics argue that xG is a trend indicator, not a deterministic predictor, and using it for single-match judgments can mislead casual audiences.
6. Ignoring Defensive Quality
xG primarily evaluates attacking events, often neglecting the defensive context. A team with low xG allowed might be executing brilliant defensive organization, which traditional xG doesn’t always highlight. Conversely, conceding goals from high-xG chances could be seen as “bad luck” when it might reflect poor defensive positioning or tactical errors.
Some advanced models attempt to include xGA (expected goals against), but even then, the interplay of tactics, pressing, and goalkeeper skill complicates the interpretation. Critics highlight that xG, in isolation, cannot fully capture the defensive side of football.
Conclusion: xG is Powerful, but Not Perfect
Expected Goals is undoubtedly a revolutionary tool in football analysis. It allows teams, analysts, and fans to measure chance quality objectively and identify underperforming or overperforming teams. However, critics rightly caution against overreliance.
xG is not a substitute for qualitative analysis, context, or human judgment. Its limitations—variance in small samples, differences between models, and inability to capture finishing skill or tactical nuance—mean it should complement, rather than replace, traditional scouting and analysis.
In short, xG is a lens, not the whole picture. Understanding its limitations helps fans appreciate the depth of football beyond just numbers and prevents over-simplified narratives about “luck” or “underperformance.”
Leave a Reply