pybaseball

"I'd walk through hell in a gasoline suit to play baseball." - Pete Rose (Charlie Hustle). Python scripts to pull MLB Gameday and stats data, build models, predict outcomes, make plays.

Predict what an MLB pitcher will throw next

Pybaseball Database: Using the Pybaseball Python Package to create a historical baseball database for any use.

V3.0 of my MLB Game Outcome Prediction Model

Simple pitch visualization tool that uses StatCast to generate charts for pitcher movement, velocity distribution, and pitch usage.

Reverse-engineered MLB The Show 26 rating formulas, with a Flask app that builds cards from real MLB stats. Any player, any season, 1897–2025.

Daily MLB Statcast ingestion pipeline → Snowflake. Powers HR & win prediction models in mlb_py.

Grades MLB pitch quality from physics alone on 2.14M real Statcast pitches, then turns the model into a front-office decision tool: pitch-design recommendations, undervalued-arm detection, year-over-year trends, and similarity-based scouting comps. Live dashboard + API.

Early Data & Game Analytics Report — automated Mariners & Rainiers dashboard built with pybaseball, mlb-statsapi, and nightly GitHub Actions.

MLB Statcast data visualization with pybaseball - 3 methods to draw baseball fields (spraychart, matplotlib, sportypy)

Calibrated probability models for MLB game and prop markets

Backtest of a bullpen fatigue "traffic light" on 2025 Statcast data: did relievers flagged by frozen workload and velocity rules do worse in the next 14 days? No evidence they did (wOBA against +0.0007, 95% CI -0.0050 to +0.0063). One season, null result published. EN/ES.

A python application to view baseball statistics and if the player is available in the provided Fantrax league. DISCLAIMER: This a personal project for education and research.

Rule-based screen for late-season bat speed change in 2026 Statcast bat tracking. Fastballs only, league-adjusted, FDR-controlled. No league-wide slowdown found.

Analyzing MLB umpire strike-zone consistency with Statcast pitch-level data (2015–2025): called-strike models, zone maps by count, and umpire/catcher leaderboards.

an end‑to‑end machine‑learning application that recommends the most appropriate reliever to bring into a game given the current game state

Using the Pybaseball-database to create a Points based ranking dashboard

MLB Statcast analytics playground with notebooks and team-themed PDF reports.

A baseball analytics dashboard in FastAPI + React: pybaseball data, park-adjusted bWAR/fWAR, and a custom Statcast-blended pitcher WAR (cWAR)

MLB HR & game winner prediction models built on Statcast data. XGBoost/LightGBM with barrel rate, exit velocity, wind, platoon splits, and pitcher matchup features. Daily pipeline with betting edge calculation. AUC: 0.620.