Data science is at the heart of everything we do.

That’s why we have exposed our core algorithms for other data scientists who want to cut some corners, save a couple of years of research and development, and get going with models that work, fit their needs, and are scientifically validated.
Player Ranking Model
Tailorable performance model, built on top of hundreds of features that we have developed over the years. Peer reviewed and in academic journals.
Physical Metrics Model
Average speed, top speed, reaction time, high-intensity running, distance covered, time spent in motion, and other metrics.
Transfer Valuation Model
Player valuation model, integrating performance, agents, contracts, selling clubs, demographics, marketing, and many other features. Peer reviewed and in academic journals.
Player Role Model
Advanced classification of role within a team or a game based on player characterestics and thousands of individual in-game events

Science and Tech Stack

Faithful to our principle of being practical, we keep our eyes on the stars, but we also keep our feet on the ground.

We use a combination of Machine Learning and traditional stastical techniques to build predictive models as well as analyze existing patterns. And our science works – check out our peer-reviewed publications below:
Player Rankings Paper
Player Valuation Paper