Credits

Daniel's OpenFPL paper and codebase introduced the multi-bucket prediction approach that inspired SmartPlay. His work on outcome-bucket classification and mixture-based xPts estimation shaped the core architecture of our prediction model.

HiGHS is the high-performance open-source solver that powers SmartPlay's squad optimiser. It solves the mixed-integer linear program that finds optimal transfers, captaincy, and lineup decisions across multiple gameweeks — reliably and fast.

Solio's open-source FPL solver inspired the design of our own optimiser. SmartPlay uses its own solver built from scratch — Open FPL Solver is not used in our project — but studying their MILP formulation and constraint modelling was invaluable in shaping our approach.

Danial's research on data-driven Fantasy Premier League strategy is the backbone of our optimiser. His framework for expected-points maximisation shaped how SmartPlay thinks about squad building.

Vaastav has been meticulously collecting and publishing FPL data for years. His dataset is one of the most valuable public resources in the FPL community and was instrumental in training our models.

Collin's Python library for the Understat API gave us clean, reliable access to advanced stats like xG and xA. It saved us countless hours and let us focus on what the data means rather than how to get it.

Chris maintains a mapping table that links FPL player IDs to Understat IDs. A small but essential piece of infrastructure that makes cross-source analysis possible.

Randdalf's cached archive of historical FPL API data filled gaps that would have been impossible to recover otherwise. A quietly invaluable resource for anyone building models on top of FPL.