Understanding the Impact of Expected Points in Premier League Analysis
In recent years, the integration of advanced probability models such as Expected Goals (xG), Expected Goals Conceded (xGA), and Expected Points (xP) has revolutionized the way football matches are analyzed. Fans, coaches, and analysts now look beyond final scores to uncover deeper insights about team performance and deserved outcomes. Here, we delve into how Expected Points can reshape our perspective on the Premier League table and what it reveals about the 2021/22 season up to February 21, 2022.
Exploring the Foundation: Expected Goals and Expected Points
Contemporary football analysis leverages metrics like xG to estimate the quality of scoring opportunities. xG assigns a value to each shot, considering the location, pressure, and type of attempt. For example, a spectacular long-range goal might carry a very low xG (such as 0.01), suggesting it would only be converted once every hundred attempts.
Expected Points (xP) builds directly on these xG values. By simulating thousands of versions of a match using the actual xG tallies generated by both teams, analysts estimate the likelihood of each possible outcome: win, draw, or loss. These probabilities are multiplied by the respective points (3 for a win, 1 for a draw, 0 for a loss) to generate a precise xP value for each team, reflecting what they statistically ‘deserved’ from the game.
Calculating Expected Points: The Method and Its Meaning
Expected Points offers a way to rate teams based on performance rather than actual results, accounting for both attacking potential (xG) and defensive solidity (xGA). Teams that register high xG and limit opponents’ xGA repeatedly are those that tend to rise in the xP rankings.
Here’s how xP is typically derived:
– Gather each team’s xG and xGA from a match.
– Run a probability distribution model to simulate thousands of match outcomes with those xG numbers.
– Calculate the percent likelihood of wins, draws, and losses from those simulations.
– Multiply each outcome’s probability by its respective point reward, summing to produce the xP.
This approach allows analysts to see whether teams’ league positions are justified by their actual performances or if luck, finishing, or other factors have skewed their fortunes.
Comparing Actual Premier League Standings to Expected Points
Let’s look at the Premier League table after matches played up to February 21, 2022:
| Rank | Team | GP | Goals | Conceded | Points |
|---|---|---|---|---|---|
| 1 | Manchester City | 26 | 63 | 17 | 63 |
| 2 | Liverpool | 25 | 64 | 20 | 57 |
| 3 | Chelsea | 25 | 49 | 18 | 50 |
| 4 | Manchester United | 26 | 44 | 34 | 46 |
| 5 | West Ham | 26 | 45 | 34 | 42 |
| 6 | Arsenal | 23 | 36 | 26 | 42 |
| 7 | Wolverhampton WM | 24 | 23 | 18 | 40 |
| 8 | Tottenham Hotspur | 23 | 31 | 31 | 39 |
| 9 | Brighton | 25 | 25 | 28 | 33 |
| 10 | Southampton | 25 | 32 | 37 | 32 |
| 11 | Leicester City | 23 | 37 | 43 | 27 |
| 12 | Aston Villa | 24 | 31 | 37 | 27 |
| 13 | Crystal Palace | 25 | 32 | 36 | 26 |
| 14 | Brentford | 26 | 27 | 42 | 24 |
| 15 | Leeds United | 24 | 29 | 50 | 23 |
| 16 | Everton | 23 | 28 | 40 | 22 |
| 17 | Newcastle United | 24 | 26 | 45 | 22 |
| 18 | Watford | 24 | 24 | 43 | 18 |
| 19 | Burnley | 22 | 20 | 29 | 17 |
| 20 | Norwich City | 25 | 15 | 53 | 17 |
Contrast this with the table based on Expected Points:
| Rank | Team | Actual Points | Expected Points (xP) | Difference |
|---|---|---|---|---|
| 1 | Manchester City | 63 | 60.5 | 0 |
| 2 | Liverpool | 57 | 54.2 | 0 |
| 3 | Chelsea | 50 | 45.5 | 0 |
| 4 | Manchester United | 46 | 37.3 | 0 |
| 5 | Arsenal | 42 | 36.3 | 0 |
| 6 | West Ham | 42 | 33.7 | 0 |
| 7 | Brentford | 24 | 32.4 | +7 |
| 8 | Crystal Palace | 26 | 31.1 | +5 |
| 9 | Brighton | 33 | 30.4 | 0 |
| 10 | Tottenham Hotspur | 39 | 29.6 | -2 |
| 11 | Southampton | 32 | 29.1 | -1 |
| 12 | Everton | 22 | 28.4 | +4 |
| 13 | Wolves | 40 | 28.1 | -6 |
| 14 | Aston Villa | 27 | 23.7 | -3 |
| 15 | Leicester City | 27 | 22.7 | -3 |
| 16 | Newcastle United | 22 | 22.1 | +1 |
| 17 | Leeds United | 23 | 21.3 | -2 |
| 18 | Watford | 18 | 18.6 | 0 |
| 19 | Burnley | 17 | 18.1 | 0 |
| 20 | Norwich City | 17 | 12.2 | 0 |
In this reimagined ranking, Brentford’s and Crystal Palace’s positions improve notably, reflecting that their performances have not been fairly translated into points. Wolverhampton Wanderers, by contrast, drop significantly, revealing a tendency to win points without consistently producing high-quality scoring chances.
Assessing the Relationship Between Possession and Performance
Another widely debated metric in football is possession percentage. While controlling the ball doesn’t guarantee results, a higher share of possession often correlates with more goal-scoring opportunities. Examining Premier League teams ranked by possession offers further perspective.
| Rank | Team | Possession (%) | Chances Created | Goals Scored | Conversion Rate |
|---|---|---|---|---|---|
| 1 | Manchester City | 67% | 227 | 63 | 28% |
| 2 | Liverpool | 61% | 227 | 64 | 28% |
| 3 | Brighton | 57% | 134 | 25 | 19% |
| 4 | Chelsea | 56% | 180 | 49 | 27% |
| 5 | Leeds United | 54% | 112 | 29 | 26% |
| 6 | Manchester United | 52% | 191 | 44 | 23% |
| 7 | Arsenal | 52% | 141 | 36 | 26% |
| 8 | Crystal Palace | 51% | 131 | 32 | 24% |
| 9 | Tottenham Hotspur | 49% | 140 | 31 | 22% |
| 10 | Southampton | 49% | 124 | 32 | 26% |
| 11 | West Ham | 48% | 168 | 45 | 27% |
| 12 | Leicester City | 48% | 129 | 37 | 29% |
| 13 | Wolves | 47% | 96 | 23 | 24% |
| 14 | Aston Villa | 46% | 106 | 31 | 29% |
| 15 | Brentford | 45% | 127 | 27 | 21% |
| 16 | Norwich City | 44% | 91 | 15 | 16% |
| 17 | Everton | 43% | 121 | 28 | 23% |
| 18 | Watford | 43% | 117 | 24 | 21% |
| 19 | Newcastle United | 41% | 100 | 26 | 26% |
| 20 | Burnley | 39% | 110 | 20 | 18% |
Arranging the table by possession reveals only small shifts, suggesting possession can be an important—though not sole—indicator of performance. Leeds United and Brighton, for example, rise in the possession-based standings, but top teams like Manchester City and Liverpool remain dominant regardless of metric. This points to a multifaceted relationship between holding the ball, creating chances, and capitalizing on those opportunities.
Key Insights: What Expected Points Teach Us About the Premier League
The use of Expected Points and related statistical models provides a more nuanced narrative of team performance than raw results convey. It highlights overachievers and underperformers and prompts us to question how much of the league table is down to execution versus underlying process.
Main takeaways include:
– Traditional league standings and xP can diverge meaningfully, revealing teams that are outperforming or underperforming their statistics.
– Teams like Brentford and Crystal Palace, despite their lower real-world point totals, have produced performances deserving of higher placements.
– Possession remains an influential but not definitive metric—high possession often correlates with chances created, but doesn’t alone guarantee success.
Conclusion: The Value of Advanced Metrics in Football Analysis
Modern data analytics, particularly xG and xP, offer powerful tools for clubs, analysts, and fans to evaluate performance beyond surface-level results. While no metric tells the entire story, combining multiple viewpoints—such as actual points, xP, and possession—provides a deeper, fairer assessment of how teams are truly performing, where they are fortunate, and where their underlying play deserves greater reward. As football data becomes more sophisticated, understanding and applying these insights will only grow in relevance across all levels of the game.
