This is the start of The Predictive Playbook — my personal experiment as a chemical engineer applying machine learning to fantasy football.
🧠 Predicting Fantasy Football with Machine Learning: A New Blog Series
Every year, millions of people draft fantasy football teams based on gut instincts, highlight reels, and outdated projections. This year, I’m doing things differently.
As a PhD-trained chemical engineer and postdoctoral researcher with a passion for data, I’m combining machine learning, statistical modeling, and a bit of football obsession to see how well we can actually predict fantasy performance using real data — not just hunches.
🎯 What This Series Is About
This blog series will document my ongoing journey building, refining, and evaluating machine learning models that predict NFL player fantasy points. The goal? To generate more accurate forecasts for drafts, weekly matchups, and even betting — all backed by open data and transparent modeling.
Here’s what to expect:
- 📅 Preseason Focus
I’m starting with predictions for the 2025 season using multi-year historical data. I’ll explore different models (like Ridge Regression, Random Forests, and XGBoost), compare their accuracy, and highlight which features seem most important for predicting fantasy success. - 📈 In-Season Updates
Once the season kicks off, I’ll shift to weekly predictions. Expect frequent model updates based on the latest player stats, injuries, and performance trends. - 💸 Sports Betting Integration
I’ll also be testing out how machine learning can support prop bets and fantasy pick’em platforms — exploring the edge that data can offer in a space dominated by gut feelings. - 🔧 Open Codebase
All my models and code will be shared on GitHub for transparency and collaboration. Whether you’re into Python, fantasy football, or just want to see how ML works in practice, you’re welcome to follow along.
🏁 Why I’m Doing This
I love exploring complex systems, and few systems are as dynamic, messy, and exciting as NFL fantasy football. This project lets me combine technical skills from my academic career with real-world predictive challenges — all while having some fun and maybe even gaining an edge.
📬 Stay Tuned
This is just the beginning. In the next post, I’ll walk through my initial model setup and share predictions for the top fantasy players of 2025 based on historical data.
Follow along as I apply machine learning to the fantasy world — and let’s see how far the data can take us.
—
Want to follow the code?
➡️ GitHub: Fantasy Football ML Project