Research
My research sits at the intersection of predictive modeling, human decision-making, and complex systems. I’m interested in how people and institutions respond to incentives, how advantages are created and eroded over time, and what good statistical models can (and can’t) say about those processes.
Labor Economics
At the Burning Glass Institute, I work on labor-market analytics and the Labor Matters newsletter, which tracks trends in employment, education, and automation. I analyze large-scale professional datasets and use text analytics to understand how opportunity, skills, and “societal good” show up across industries and educational backgrounds.
Methods: SQL querying, text and embedding features, regularized regression, clustering, Elo-style models
Sports Analytics
Through my basketball analytics work and my blog, I study how teams create, trade, and defend advantages over the course of a possession and a season. I’ve developed defensive versatility and hustle metrics, as well as predictive models for rookie-scale contracts, shooting, and player development, often blending tracking-style information with play-by-play data.
Some of this research has been presented at Yale University to an audience of statistics faculty, front-office executives, and graduate students.
Methods: regression variants, hierarchical / mixed-effects models, clustering, Elo-style ratings, Monte Carlo simulations, computer vision, geospatial data