Our Definition of Expected Points (xP)

In football analytics, Expected Points (xP) has emerged as a key performance indicator (KPI), reflecting the likelihood of a team winning a match based on the quality and quantity of chances created. This metric has earned attention within our team and across the broader analytical community. But how is xP calculated, and what makes it so insightful?

The Essence of Expected Points

xP aims to quantify the probability of winning if a game were replayed multiple times, effectively minimizing the randomness that a single match can introduce. For instance, a team might secure a win by converting a lone chance, while their opponent fails to score despite numerous opportunities. Over a series of replays, the outcome would likely favor the initially unsuccessful team, highlighting the underlying performance levels xP seeks to capture.

Redefining xP: Beyond Shots and Goals

Traditionally, xP calculations have focused on expected goals (xG), a metric assessing the quality of scoring opportunities. While xG offers profound insights into game dominance, it overlooks aspects like possession and the generation of threats (xT), which also contribute to a team's performance.

Expanding the xP Model

Critiques of xG often point out its narrow focus on shots, neglecting how teams dominate matches through possession or threat generation. However, correlations exist between points and these overlooked metrics: 0.54 for possession and 0.5 for xT, suggesting their significance. This realization prompts a critical question: Why not incorporate these KPIs into the xP calculation?

A Comprehensive Approach to Performance Metrics

Embracing a broader perspective, we have developed our own expanded xP model that incorporates possession and xT alongside xG. This approach aims to capture a more accurate reflection of a team's performance, acknowledging the multifaceted nature of football dynamics. Our analysis reveals that including these additional indicators not only enriches the model but also improves its descriptive power.

Towards a More Informed Understanding of Team Performance

This expanded xP model signifies our commitment to refining football analytics, offering a nuanced understanding that surpasses conventional metrics. By integrating possession and xT, we venture closer to depicting the true essence of a team's performance, paving the way for more strategic insights and decision-making.

Conclusion

The evolution of our xP model, from a sole reliance on xG to incorporating broader performance indicators, marks a significant advancement in football analytics. This comprehensive approach ensures a deeper and more accurate analysis of what contributes to a team's success, reinforcing the importance of a multifaceted perspective in evaluating performance.