Technologies: Python, Jupyter Notebooks
Skills & Methodologies: Probabaility Distributions
GitHub Repository
In collaboartion with the Baltimore Orioles I attempted to fit parametric probability distributions to bivariate batted ball data. Batted ball data consists of exit velocity and launch angle. The goal is to devlop a methodology that will fit a probabaility distribution from a selected parametric family to each player's batted ball data for a given season. Then, using this distribution, batted balls can be "sampled" from for the purposes of simulation. Ultimately, I was unable to construct a suitable probabaility distribution that statisfied the stakeholder's desired level of complexity and accuracy. My distrubtions either consisted of too many parameters or were not accurate enough. I plan to revisit this project when time allows.