Both roles involve building on top of proprietary tracking and pose data to create metrics, models, and analyses that elite sports organizations actually use. Day-to-day the work looks like this across both:
Build and transform new data sources into tables, features, and structures that are easy for the team and our clients to build on
Develop, extend, and validate models — including event-probability models and athleticism models — ensuring data representation supports both current and future use cases
Build metrics and analyses that NHL and NBA clients rely on to make decisions, and support client-facing work by digging into the data to answer their questions directly
Extract meaningful features from high-dimensional tracking and pose data, and update existing models to incorporate new signals as they become available
Validate models and outputs — your own and others' — with enough rigor that the team can trus...