Selected Work
Worked examples behind the services: the analysis, models and engineering each one is built on.
Data Scouting
Finding players statistically: PCA style-mapping, anomaly detection and contextual scatterplots that surface names worth a closer look, each paired with a sense of how much to trust the number.
See the visuals →Data Analysis
Match and season breakdowns (buildup patterns, set-piece efficiency, territory and retention) built to hold up under scrutiny, not just look plausible.
See the visuals →Data Engineer
Model pipelines: Danger model shot-quality validation, xT and the shot-clock model as a Markov Decision Process, structured to run the same way every week.
See the visuals →Data Solutions
Critical thinking applied to design questions: testing whether a radar chart's visual story matches the data, and why the same metric can mean four different numbers depending on provider.
See the analysis →Models
Recent model work: Wins Above Replacement, Contextual xT, Expected Fails Forced and goalkeeper tracking profiles, each written up in full.
See the models →