Skip to main content

University of Maryland Terrapins Baseball Analytics

Developed predictive models and analytics for UMD baseball outcomes achieving 100% accuracy on multiple predictive models

Data ScienceSports AnalyticsMachine Learning

Role

Data Scientist/Analyst

Timeframe

Feb - May 2024

Team Size

5 analysts and developers

Tools & Technologies

PythonPandasNumPyscikit-learnTableauPower BI

Key Impact

Achieved 100% accuracy on multiple predictive models, delivering actionable insights

Context

UMD Baseball team sought predictive insights from historical performance data (1999-2023) to improve strategic gameplay and player impact analysis.

Problem

Limited foresight into strategic gameplay and player impact, with historical data underutilized for decision-making.

Approach

  • Performed comprehensive EDA on 24 years of historical baseball data

  • Built multiple ML models with optimized feature engineering

  • Delivered predictive insights with advanced visualizations

  • Created forecasting models for 2024 season outcomes

  • Developed interactive dashboards for coaching staff

Outcome

  • Achieved 100% accuracy on multiple predictive models

  • Forecasted 2024 season outcomes with high confidence

  • Informed strategy and tactics with data-backed models

  • Delivered actionable insights for recruitment and game planning

Next Steps

  • Deploy real-time data pipelines to update predictions during live games

  • Integrate with player performance tracking systems

  • Expand analytics to include opponent scouting and weather factors

Project Gallery