Stephen Akwaowo

Context-Adjusted Induced Vertical Break Metric (ΔiVB⁺)

Developed a context-adjusted four-seam fastball metric using XGBoost to estimate expected induced vertical break from release geometry, spin characteristics, velocity, and approach angle. Article includes model explanation and a searchable 2025 MLB leaderboard.

NBA Player Similarity & Archetype Engine 🚧 Under Construction

Developing a role-aware NBA player similarity tool that identifies stylistic comparisons through shot profile, usage, playmaking, efficiency, and defensive indicators. Designed to support player archetypes, evaluation, and roster-construction analysis.

MLB Batter Analysis Dashboard

Interactive Shiny application for the Clustering MLB Hitters project, featuring Gaussian Mixture Models, sequential dimensionality reduction, and the BONDS Index.

Clustering MLB Hitters Documentation

Technical documentation covering Gaussian Mixture Models, sequential dimensionality reduction, the BONDS Index, and the development of the accompanying Shiny application.

2022 NBA MVP Predictor

Machine-learning project comparing XGBoost, linear regression, and random forest models to predict NBA MVP voting outcomes.

GitHub

Substack

Tableau

Medium