The Modeled Scouting Backend API is a statistical framework for analyzing and predicting performance in FIRST Tech Challenge (FTC) teams and competitions.
This project is heavily inspired by Statbotics — an outstanding open-source resource and model for FRC created by Av Gupta.
Much of our approach, structure, and modeling philosophy is adapted from Statbotics’ EPA (Estimated Performance Average) system. While it has been previously more common to see the use of OPR/ELO and Average Scores, this is far less accurate than EPA. You can read their blog post on it here.
We’ve reworked these concepts to fit the unique dynamics of FTC, including smaller alliances, different scoring systems, and more limited data availability.
Note: This is not an official Statbotics project. It is an independent adaptation designed to bring similar analytical insights to FTC teams for scouting, performance tracking, and strategy.
While Statbotics uses CockroachDB, this project adopts a Cloudflare D1 for faster API performance and simpler scalability.
You can explore the original Statbotics project and documentation here:
👉 https://statbotics.io
This model is part of a larger open resource that will be released later this FTC season — stay tuned for more details!
© 2025 Hivemind Robotics
Licensed under the MIT License