pyTSPA-toolbox 0.1.1

A toolbox for analyzing and visualizing team sports performance

pyTSPA - Toolbox for Analyzing and Visualizing Team Sports Performance

A toolbox for analyzing and visualizing team sports performance, with a particular focus on football match data.
The toolbox includes modules for data processing, statistical analysis, match outcome prediction, and result visualization.

---

Current Version: 0.1.1

Planned Release: May 2025

---

Introduction

pyTSPA was developed to support coaches, analysts, and sports scientists in extracting actionable insights from football match data.
Designed with a football-centric approach, the toolbox facilitates data handling, statistical analysis, match outcome prediction using machine learning, and comprehensive data visualization.

---

Description

The toolbox is divided into the following modules:

• Data Handling:

  • Data input/output, cleaning, and profiling (CSV, Excel).

• Metrics:

  • Calculation of match and team statistics (Win Percentage, Pythagorean Expectation)
  • Predictive models using logistic regression for match outcome prediction.

• Visualization:

  • Graphical representation of match data, statistical summaries, and prediction results using bar charts, scatter plots, and pie charts.

---

For more information, visit the official documentation.

Installation

From PyPi

Install pyTSPA using pip

pip install pyTSPA-toolbox

From source

Clone the git repository

git clone https://github.com/vargaheni05/pyTSPA-toolbox.git

Change directory to the cloned repository

cd pyTSPA-toolbox

Install with pip

pip install .

Building the documentation

Install required packages

pip install -r docs/requirements.txt

Call Sphinx build command

sphinx-build -M html docs/source docs/build

On Windows you can also run the make.bat file

.\docs\make.bat html

The documentation should be available in the docs/build directory as html files
This includes the example codes as tutorials

Correspondence

Henrietta Varga ([e-mail])
Marcell Szögi ([e-mail])
Benedek Kardos ([e-mail])