How Much Does Home Field Advantage Matter in Soccer Games? A Causal Inference Approach for English Premier League Analysis

Katherine Price, Hengrui Cai, Weining Shen, Guanyu Hu, Cai, Zhanrui, Hu, GuanyuarXiv 2022 · DOI · arXiv

Abstract

In many sports, it is commonly believed that the home team has an advantage over the visiting team, known as the home field advantage. Yet its causal effect on team performance is largely unknown. In this paper, we propose a novel causal inference approach to study the causal effect of home field advantage in the English Premier League. We develop a hierarchical causal model and show that both league-level and team-level causal effects are identifiable and can be conveniently estimated. We further develop an inference procedure for the proposed estimators and demonstrate its excellent numerical performance via simulation studies. We implement our method on the 2020-21 English Premier League data and assess the causal effect of home advantage on eleven summary statistics that measure offensive and defensive performance and referee-assessed disciplinary outcomes. We find that the home field advantage resides more heavily in offensive statistics than it does in defensive or referee statistics. We also find evidence that teams that had lower rankings retain a higher home field advantage.

Introduction

While sports fans and analysts intuitively understand that playing at home provides an edge, quantifying this "Home Field Advantage" (HFA) as a causal effect is a significant statistical challenge. Most existing analyses rely on simple associations—such as win rates or goal differentials—which are inherently confounded by team quality and the structural nature of league scheduling.

To move beyond correlation, this research introduces a hierarchical causal inference framework designed to isolate the venue effect from the intrinsic skill of the teams involved. By applying this model to the English Premier League (EPL), we can determine not just if HFA exists, but where it manifests in the flow of a game.

The Challenge of Causal Inference in Sports

Estimating the causal effect of a venue presents two unique hurdles for standard statistical methods:

  1. Absence of a Control Group: In a professional league, there are no "neutral" baseline games. Every match is either home or away, making it difficult to establish what a team's performance would look like without the venue treatment.
  2. Interference and Anti-correlation: A team’s status as the "home" team is perfectly anti-correlated with their opponent’s status as the "away" team. Because the treatment is not independently assigned to each unit, standard propensity score matching fails.

Methodology: The Pairing Strategy

To overcome these hurdles, the study proposes a hierarchical causal model based on the potential outcomes framework. The core innovation lies in a pairing strategy. By analyzing the two games played between any two teams (Team $i$ and Team $j$) within a single season—one at each team's stadium—the researchers can mathematically decompose the match outcomes.

By subtracting the results of these paired games, the "intrinsic skill" variables (nuisance parameters) cancel out, leaving a linear system where the remaining variables represent the HFA effects. The model treats individual team HFA as a random effect drawn from a league-wide distribution, allowing for the estimation of both a general league effect ($\Delta$) and team-specific deviations.

Data and Metrics

The model was tested using data from the 2020-2021 EPL season (provided by Hudl & Wyscout), covering 380 matches. The researchers analyzed 11 specific performance metrics categorized into three domains:

  • Offense: Attacks with Shot, Reaching Opponent Box/Half, Shots from Box/Danger Zone, Successful Key Passes, Touches in Box, and Expected Goals (xG).
  • Defense: Defensive Interceptions and Shots Blocked.
  • Refereeing: Yellow Cards.

Key Findings

HFA as an Offensive Phenomenon

The analysis reveals that HFA is primarily an offensive boost. The causal effect was statistically significant across nearly all offensive metrics. For instance, playing at home resulted in a league-wide average increase of 3.592 instances of "Reaching Opponent Half" per game ($p < 0.001$) and 1.568 additional "Attacks with Shot" ($p = 0.005$). Conversely, defensive metrics like interceptions and blocked shots showed no significant causal link to the venue.

The Underdog Advantage

A striking insight from the team-level analysis is that HFA is not distributed equally. Lower-ranked teams (those in the bottom 50% of the standings, such as Fulham and Brighton) benefited significantly more from playing at home than the league's elite. While top-tier teams like Manchester City demonstrated "venue-invariance"—performing at a high level regardless of location—underdogs appeared to rely on the home environment to bridge the talent gap.

Debunking Referee Bias

Contrary to the popular narrative that home crowds pressure referees into favorable decisions, the causal analysis found no significant evidence of referee bias during the 2020-21 season. The causal effect on "Yellow Cards" was near zero (-0.026, $p = 0.834$), suggesting that disciplinary outcomes are driven by gameplay rather than venue.

Conclusion and Future Directions

This research provides a robust framework for isolating the causal impact of home field advantage, demonstrating that the effect is most pronounced in a team's offensive creativity and pressure.

However, the 2020-21 season was unique due to the COVID-19 pandemic, which resulted in many "ghost games" played in empty stadiums. Future work applying this model to seasons with full crowds could help isolate the specific contribution of crowd noise versus other factors like travel fatigue. Additionally, extending the model to include mediation analysis could help researchers understand the exact pathway from increased offensive pressure to final match scores.

Summary updated

Abstract

In many sports, it is commonly believed that the home team has an advantage over the visiting team, known as the home field advantage. Yet its causal effect on team performance is largely unknown. In this paper, we propose a novel causal inference approach to study the causal effect of home field advantage in English Premier League. We develop a hierarchical causal model and show that both league level and team level causal effects are identifiable and can be conveniently estimated. We further develop an inference procedure for the proposed estimators and demonstrate its excellent numerical performance via simulation studies. We implement our method on the 2020-21 English Premier League data and assess the causal effect of home advantage on eleven summary statistics that measure the offensive and defensive performance and referee-assessed disciplinary outcomes. We find that the home field advantage resides more heavily in offensive statistics than it does in defensive or referee statistics. We also find evidence that teams that had lower rankings retain a higher home field advantage.
BibTeX
@misc{price2022how,
  author = {Katherine Price and Hengrui Cai and Weining Shen and Guanyu Hu and Cai, Zhanrui and Hu, Guanyu},
  title = {{How Much Does Home Field Advantage Matter in Soccer Games? A Causal Inference Approach for English Premier League Analysis}},
  year = {2022},
  eprint = {2205.07193},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2205.07193},
}
Figure