We work with professional sports organisations — currently eight clubs across the NBA and MLS — to turn tracking data, event data, and video analysis into decisions: player recruitment rankings, injury risk assessments, in-game tactical adjustments, and season performance projections. The data we work with is genuinely interesting: sub-second spatial tracking from Second Spectrum systems, event-by-event match logs, biometric monitoring data from training sessions, and historical contract and performance archives going back fifteen years. The modelling is interesting too — survival analysis for injury modelling, graph-based approaches for passing pattern analysis, Bayesian hierarchical models for projecting performance under uncertainty. We're looking for a data scientist who has thought carefully about the specific challenges of small-sample, high-variance sports data: where you have 38 games in a season, not 38,000 observations, and where the signal is real but easily confounded by team effects, opposition strength, and a dozen other factors that naive models miss. Strong statistical reasoning, not just ML tooling, is the priority.
Responsibilities
Build and validate player performance projection models for recruitment analysis with appropriate uncertainty quantification
Develop injury risk models using biometric and training load data with clinically-meaningful output formats
Analyse passing and positioning data for tactical insights using spatial and graph-based approaches
Present model outputs, uncertainty ranges, and recommendations to club coaching and management staff in accessible formats
Maintain and document the modelling pipeline and evaluation benchmarks used across all club engagements
Requirements
3–5 years of data science experience — sports analytics preferred, but deep statistical rigour in any high-variance domain is valued
Python — Pandas, NumPy, Scikit-learn, and at least one deep learning framework
Strong applied statistics: Bayesian inference, survival analysis, mixed effects models — you've implemented and interpreted these, not just referenced them
SQL for extracting and preparing large structured datasets
Strong data visualisation skills for presenting findings to non-technical coaching and management staff
Experience with spatial or graph-based data analysis is a meaningful plus
R fluency is useful given the breadth of statistical modelling we cover
Benefits
Serious statistical modelling work in a domain most data scientists find genuinely exciting — and with access to data most cannot get near
Full remote with occasional travel to client club facilities (4–6 times per year)
$90,000 – $112,000 base salary + performance bonus
$1,500 annual conference budget
Access to some of the richest sports data sets available outside top-tier clubs' internal research teams