2025 · Causal Inference
English Premier League Spending Analysis
Causal Inference Sports Analytics R
English Premier League Spending Analysis
University of Maryland | Jan 2025 – Apr 2025
R Synthetic Control Difference-in-Differences ggplot2
Project Overview
Applied quasi-experimental causal inference methods to examine the relationship between squad investment and team performance in the English Premier League.
Key Contributions
Causal Framework: Applied synthetic control and difference-in-differences (DiD) methods to estimate causal effects of squad investment on league position, controlling for confounding factors
Model Implementation: Constructed synthetic counterfactuals for treated clubs using convex combinations of donor pool teams, validating pre-treatment parallel trends
Statistical Inference: Conducted placebo tests and permutation inference to assess statistical significance, finding significant ROI heterogeneity across club tiers (top-6 vs. mid-table)
Technologies Used
- Languages: R
- Methods: Synthetic Control, Difference-in-Differences
- Visualization: ggplot2
- Data Sources: TransferMarkt, Premier League
Key Findings
| Analysis | Result |
|---|---|
| Treatment Effect (Top-6) | Significant positive |
| Treatment Effect (Mid-table) | Heterogeneous |
| Pre-treatment Parallel Trends | Validated |
| Placebo Tests | Passed |