Disclaimer: The following article is an educational case study. All player names, team names, statistics, and scenarios are fictional and created for illustrative purposes only. No real-world matches, players, or data are referenced.
Expected Goals (xG) Season Review: Deconstructing the 2023/24 Campaign Through Player & Team Statistics
The 2023/24 season in European football offered a rich dataset for analysts to test the explanatory power of Expected Goals (xG). While the final league tables told one story, the underlying xG metrics often revealed a different, more nuanced narrative. This case study examines how a season review, grounded in player and team statistics, can move beyond simple win-loss records to diagnose performance, identify systemic issues, and predict future regression or progression.
The Opening Statement: xG as a Diagnostic Tool, Not a Prophecy
The core premise of xG is that not all shots are created equal. A 30-yard speculative effort carries a lower probability of scoring (e.g., 0.02 xG) than a close-range header from a cutback (e.g., 0.45 xG). By aggregating these probabilities, xG provides an estimate of the number of goals a team should have scored or conceded, based on the quality of chances created and allowed. This season review focuses on two primary metrics: xG For (quality of chances created) and xG Against (quality of chances conceded). The difference, xG Delta (xG For – xG Against), is often a more reliable indicator of sustainable performance than actual goal difference.
Phase 1: The Overachievers and Underachievers – Where xG Reveals a Mirage
A classic application of xG in a season review is identifying teams and players who significantly outperformed or underperformed their expected numbers. Consider a fictional Premier League team, "Riverside United."
| Metric | Riverside United (Actual) | Riverside United (xG) | Difference (G – xG) |
|---|---|---|---|
| Goals Scored | 68 | 54.2 | +13.8 |
| Goals Conceded | 42 | 51.8 | -9.8 |
| Points | 72 | ~58 (estimated) | +14 |
The Analysis: Riverside United finished 3rd in the league. However, their xG data suggests they were a mid-table side in terms of chance creation and prevention. The +13.8 goal overperformance is a massive outlier, likely driven by exceptional finishing from a single player (e.g., a striker converting 25 goals from 18.5 xG) and unsustainable goalkeeping heroics. An analyst would flag this team as a prime candidate for regression the following season. The club’s director of football, relying on this analysis, might decide against paying a premium for the striker, suspecting his finishing rate is unsustainable.
Phase 2: Player-Level Decomposition – The Individual xG Story
Breaking down xG to the player level provides granular insight. For instance, a winger with a high xG but low actual goals might be creating excellent chances for himself but lacking composure. Conversely, a midfielder with a high xA (Expected Assists) but few actual assists is likely creating high-quality chances for teammates who are underperforming.
The Case of a "False Positive": A forward named "Marco Silva" (fictional) scored 12 goals from a total xG of 8.5. This is a significant overperformance. However, a deeper look reveals that 6 of his goals came from penalties (xG per penalty ~0.76). Excluding penalties, his open-play xG was 6.2, and he scored 6 open-play goals. This suggests his finishing was not exceptional; rather, his goal tally was inflated by penalty duties. An analyst would adjust his valuation downward, noting that his non-penalty xG was merely average.
Phase 3: Systemic and Tactical Correlations – Linking xG to Formation and Pressing xG data becomes most powerful when correlated with tactical systems. This season review examines how different formations influence xG generation and suppression.
Formation vs. xG Metrics (Fictional League Data)
| Formation | Avg. xG For per 90 | Avg. xG Against per 90 | Avg. PPDA (Passes per Defensive Action) | Typical Weakness |
|---|---|---|---|---|
| 4-3-3 | 1.65 | 1.12 | 9.2 (High Press) | Vulnerable to counter-attacks through the half-spaces if the lone pivot is bypassed. |
| 4-2-3-1 | 1.48 | 1.05 | 10.8 (Mid-Block) | Risk of creating low-quality chances from outside the box if the No.10 is marked out. |
| 3-5-2 | 1.32 | 1.25 | 12.5 (Deep Block) | Struggles to create high-xG chances in the box due to lack of width in the final third. |
Interpretation:
- 4-3-3 System: Teams using this formation often generated a high xG For due to wingers cutting inside. However, their high pressing (low PPDA) meant they were susceptible to quick transitions, leading to a higher xG Against than a 4-2-3-1.
- 4-2-3-1 System: This setup offered a balance. The double pivot provided defensive solidity (low xG Against), but the solitary striker could sometimes be isolated, leading to a slightly lower xG For.
- 3-5-2 System: The data suggests this formation was a defensive trade-off. While solid in central areas, it conceded a higher xG Against than expected, likely from crosses and set pieces, and struggled to generate high-quality chances (low xG For).
Phase 4: The Contextual Limitations – Why xG is Not the Final Word
A rigorous season review must also acknowledge the model’s limitations. xG does not account for:
- Goalkeeper Quality: A world-class save on a 0.45 xG shot is not captured; the model still assigns 0.45 xG against the team.
- Defensive Pressure: A shot from 12 yards with a defender closing down is treated similarly to one with no pressure.
- Game State: Teams trailing often take more, lower-quality shots, inflating their xG without reflecting true dominance.
- Set Pieces: xG models for corners and free-kicks are less mature and often underweight the chaos of a crowded box.
Conclusion: A Summary Table for the Season Review
The 2023/24 season review, when viewed through the lens of xG, provides a powerful corrective to the "results-based analysis" that dominates mainstream coverage. The table below summarizes the key diagnostic insights.
| Type of Team/Player | xG Signature | Interpretation | Recommended Action |
|---|---|---|---|
| Overachieving Team | High Goals, Low xG | Unsustainable finishing/Goalkeeping | Sell high on key players; expect regression. |
| Underachieving Team | Low Goals, High xG | Poor finishing or bad luck | Buy low on finishers; expect positive regression. |
| High xG, Low xA Player | Good shot volume, poor creation for others | Selfish or isolated attacker | Assess tactical fit; may need a partner. |
| Low xG, High Goals Player | Finishing outlier | Unsustainable hot streak | Do not overpay based on goal tally alone. |
For a deeper dive into how these metrics are used in real-time analysis, see our guide on analyzing xG in match reports. To understand how chance creation is measured, compare xG with expected assists (xA). Finally, remember that xG is only one piece of the puzzle; it must be contextualized with possession percentage and outcome to form a complete tactical picture. The best analysts use these tools not to predict the future, but to ask better questions about the present.