Sports Analytics

Sports Analytics

The same statistical rigour I apply to concept drift and non-stationary data streams translates naturally to sports. Here I explore performance metrics, player analytics, and predictive modelling in football, basketball, and beyond.

Football Python

FIFA World Cup 2022 Analysis

Shot maps, xG analysis, passing networks, pressing intensity (PPDA), and player performance radars using StatsBomb open data.

View on GitHub โ†’
Basketball R

Player Efficiency Under Pressure

Analysing how player performance metrics shift in high-leverage game situations โ€” a sports application of non-stationarity and distribution drift.

๐Ÿ“ Coming soon on GitHub

Football Machine Learning

Transfer Value Prediction

Predicting player transfer market values using performance data, positional metrics, and age-progression curves.

๐Ÿ“ Coming soon on GitHub

Multi-Sport Stats

Drift in Team Performance

Applying concept drift detection algorithms to time series of team performance indicators across a season โ€” when does a team fundamentally change?

๐Ÿ“ Research in progress

Analysis Posts

Detailed analyses and tutorials will be posted here. Topics will include:

  • How to build a shot map in Python with mplsoccer
  • Applying the KS-test to detect structural breaks in team form
  • Clustering playing styles with k-means on event data
  • A Ghanaian Premier League data analysis

Subscribe via RSS or check back soon.

Data Sources & Tools I Use

StatsBomb Open Data FBref Understat Basketball-Reference mplsoccer pandas scikit-learn matplotlib ggplot2 tidyverse
See My GitHub for Code โ†’