Michael Pilger · Quantitative Portfolio
Data analysis, statistical modeling, and applied research.
Selected work in forecasting, simulation, machine learning, sports analytics, and data-product development. The principal research project is Window-Adjusted WAR, a team-specific framework for prospective MLB player value.
About
This section is reserved for a short biography and professional background.
Biography
- Education
- Experience
- Research interests
- Contact
Research and analytical projects
Projects are organized around the question being studied, the methods used, and the resulting analysis or application.
Window-Adjusted WAR (wWAR)
A prospective MLB player-value framework that holds a player's projected WAR fixed and estimates how efficiently those wins convert into championships for each franchise over a five-year competitive window.
League Analytics Platform
A live statistics and league-operations platform covering standings, schedules, player records, awards, projections, data ingestion, and administrative workflows.
XGBoost Classification
A supervised-classification study using the UCI Adult dataset, with logistic-loss optimization, hyperparameter tuning, regularization, and explicit checks for generalization.
Voice Reception & Workflow Systems
Small-business call-handling and workflow prototypes using structured routing, API integrations, SMS actions, and GoHighLevel configuration.
Technical methods represented in the portfolio
The categories below summarize methods used across the research, modeling, and software projects.
Statistical modeling
- Regression and classification
- Monte Carlo simulation
- Forecasting and counterfactual design
- GAM, EFA, CFA, and SEM
Python
- pandas and NumPy
- scikit-learn and XGBoost
- Model evaluation
- Data preparation and pipelines
R and inference
- Latent-variable modeling
- Validation and diagnostics
- Reproducible analysis
- Statistical reporting
Data systems
- SQL-backed web applications
- Cloudflare deployment
- API integrations
- Analytical interfaces
The wWAR project combines a defined research question, prospective forecasting, Monte Carlo simulation, temporal validation, reproducible outputs, and a public analytical presentation.