A roundup of tutorials, courses, and books that use Fantasy (American) Football data for their subject.
2024 update: In this article I share how fantasy football is the ideal dataset to use when you’re learning new technologies. I’m putting this into practice in a book I am writing 📕 with O’Reilly! It’s called Hands-On APIs for AI and Data Science, and the early release is out now — take a look!
Step 1: Starting With Toy Data
When learning a new technology or data science technique, the first step is often following along with a demonstration project or tutorial. These often use toy data such as Iris, Titanic, Penguins, or (for Microsoft fans) the classic Northwind. These datasets are typically much cleaner than datasets found in the wild, so the learner can focus on the technique or tool they are learning. But clean data can only teach you so much.
Step 2: Pick Real World Data
The next level of learning is to build products that you enjoy and understand. In the Data Science field, we give a lot of attention to the importance of Domain Expertise, another term for business knowledge. Building products with data you understand helps in several key ways:
- Solving a real-world problem pushes you into an active-learning mode.
- You’re more skeptical of pat answers and broad claims in an area where you know an exception to every rule.
My Domain Of Choice: Fantasy Football
Fantasy (American) Football has often been a convenient data domain for me. The NFL games and fantasy websites generate quite a bit of tabular data, so it’s handy for a variety of data science and analytics methods. There are also some good opportunities for working on API, web, and mobile development. It’s a long-time hobby of mine, so I am never short on questions I’d like to ask the data about next week and next year. (Alas, it doesn’t confess easily, no matter how much it is tortured.)
Here is roundup of courses, tutorials, books, and articles I have found that use Fantasy Football as their domain:
Learn to Code APIs With Fantasy Football
Hands-on APIs for AI and Data Science — I am publishing a book with O’Reilly in 2025 that uses fantasy football to demonstrate building APIs and using them in data science and generative AI. You’ll also build a professional-quality portfolio project. Check it out!
Learn Entrepreneurship From Fantasy Football
Business Lessons From Fantasy Football — Jack Plotkin shares parallels between fantasy football and running a business including hiring and business strategy. Interesting stuff and quite thorough.
Learn Python, R, and Data Science With Fantasy Football
Football Analytics with Python and R — I was a tech reviewer of this book by Eric Eager and Richard Erickson, and it’s highly recommended. The authors build step-by-step from basic to advanced modeling and visualizations using both R and Python using NFL data.
Fantasy Football Data Pros — I have taken this paid course that uses Fantasy Football to demonstrate a variety of Python and data science techniques. The instructor Ben Dominguez includes a good combination of videos and step-by-step lessons. When I took the course, it included some fun extras including a Discord server and even a fantasy league including other students. (Side note: there are quite a few free lessons on the blog of this site as well.)
Learn to Code with Fantasy Football — This is an eBook by Nathan Braun that looks good, but I haven’t had a chance to read it yet. The author also has a developer kit available with some add-ons.
Here are a few more articles that aren’t technically ‘teaching’ the techniques but have very nice demonstrations by people generously sharing their learnings:
- Using Machine Learning to Predict Fantasy Football Points by Jim King, 2020
- Taking Fantasy Football Analytics to the Next Level with Automated Machine Learning by Rajiv Shah, 2018
Special credit is due to the maintainers of the nflfastR package that many of these resources use. Keep up the good work!
Learn Linear Programming
Using Python and Linear Programming to Optimize Fantasy Football Picks — I was not familiar with Linear Programming before reading this excellent article by Branko Blagojevic. He explains the types of problems that Linear Programming is intended for, then demonstrates the methods using daily fantasy. It definitely has whetted my appetite to learn more about this unique discipline, which is distinct from machine learning but can be applied to some similar problems.
Learn Java Script
Building A Fantasy Football App with JavaScript Objects — Ethan Jarrell demonstrates some simple JavaScript with Fantasy Football. He also includes some really slick-looking sketches for the app design.
Learn PowerShell Programming
Learn How To Code With Football — I can’t vouch for this one, but the Udemy description says it covers PowerShell Scripting and API calls.
Learn about Luck and Monte Carlo Analysis
Luck and the Law: Quantifying Chance in Fantasy Sports and Other Contests — This is a 2018 research paper that uses Monte Carlo Analysis to determine where daily fantasy football games fall on the luck-to-skill spectrum. (This apparently had some implications for the legal status of daily fantasy at the time.) Fantasy Football regulars may not agree with some choices in their experiment design, but that’s what makes it a useful way to think of applying statistical methods to an area you know well.
What Am I Missing?
I am sure there are some great resources that I haven’t included. Feel free to comment below and pass those along.
