Projects

Project Federer — A data-driven tennis analytics system that encoded every shot in a match across five dimensions, yielding 1,326 possible shot types, then ranked shot effectiveness using a custom “Benefit Index” metric. Built from a recorded match, a Python search client, and applied to real match data, the project demonstrated that granular shot-level analysis can surface strategic insights invisible to traditional statistics.

Project Insight — A probabilistic model that derives closed-form equations for the probability of winning a tennis game, set, and match from a single input: the probability of winning an individual point. The model uses infinite geometric series and combinatorics to handle tennis’s layered scoring structure, and was validated against real match data, revealing how tennis’s hierarchical structure dramatically amplifies small edges in point-winning ability.