The newest release of the 𝒖𝒏𝒓𝒂𝒗𝒆𝒍𝒔𝒑𝒐𝒓𝒕𝒔 Python package (v1.1.0) now includes the functionality of my latest research paper titled: β€œπΈπΉπ‘ƒπΌ: πΈπ‘™π‘Žπ‘ π‘‘π‘–π‘ πΉπ‘œπ‘Ÿπ‘šπ‘Žπ‘‘π‘–π‘œπ‘› π‘Žπ‘›π‘‘ π‘ƒπ‘œπ‘ π‘–π‘‘π‘–π‘œπ‘› πΌπ‘‘π‘’π‘›π‘‘π‘–π‘“π‘–π‘π‘Žπ‘‘π‘–π‘œπ‘› 𝑖𝑛 πΉπ‘œπ‘œπ‘‘π‘π‘Žπ‘™π‘™ (π‘†π‘œπ‘π‘π‘’π‘Ÿ) 𝑒𝑠𝑖𝑛𝑔 π‘‡π‘’π‘šπ‘π‘™π‘Žπ‘‘π‘’ π‘€π‘Žπ‘‘π‘β„Žπ‘–π‘›π‘” π‘Žπ‘›π‘‘ πΏπ‘–π‘›π‘’π‘Žπ‘Ÿ π΄π‘ π‘ π‘–π‘”π‘›π‘šπ‘’π‘›π‘‘β€.

The aim of π„π…ππˆ is to easily assign a formation and associated position labels to segments of tracking data. These segments can be individual frames of a full game, possessions, periods or any (custom) time interval you want (e.g. every 5 minutes). π„π…ππˆ computes the cost a fitting a set of formation templates (from the π’Žπ’‘π’π’”π’π’„π’„π’†π’“ library) to scaled player positions, it picks the formation template with the lowest assignment cost using linear sum assignment and allocates each player their position label. By releasing EFPI as open-source software, I hope to provide a solid solution for this rudimentary, and commonly needed task.

More information on this Python package can be found in the GitHub repository linked below.

Readme Card

EFPI Example

<
Previous Post
πŸŽ™οΈ GraphEPV at PyData London 2025
>
Blog Archive
Archive of all previous blog posts