One independent analyst, public play-by-play data, and every method shown in full
I'm C. B. Zakarian. NFL Analytics Textbook is one person's attempt to write down, in one place, how modern football analysis actually works.
The site started as a folder of R and Python scripts I kept rerunning to settle arguments with myself - is fourth-and-short really a go, is that quarterback good or just playing behind a good line, why does the broadcast keep quoting a stat that falls apart the moment you pull the play-by-play. The notes eventually got long enough to be chapters, the chapters became a textbook, and the one-off questions became explainers and calculators.
There is no staff and no "we" here. It's one person who likes the math and would rather show you the calculation than ask you to trust a take. If you're a fan trying to follow the analytics conversation, a fantasy player, an aspiring analyst, or a bettor who wants to understand the numbers - the material is written for you, and all of it is free.
Everything here is built from public data - above all the nflverse play-by-play releases (nflfastR and nfl_data_py) that anyone can download for free. I show the math: the formula, the code, and the caveats travel with every metric, so you can rerun the numbers instead of taking my word. The bundled dataset behind the explainers' exhibits - the nflverse games file with every result since 1999 - is served openly at /data/ (see SOURCE.txt for provenance), so you can check any figure against the same file I used. And the site keeps one pledge on every page: no invented statistics. If a number appears here, it came from a real data pull or a named source. When the data can't answer a question, the article says so instead of guessing.
Every chart and table starts from data you can download yourself - no proprietary feeds, no numbers you can't check.
Concepts come with runnable R and Python code. Copy it, modify it, and check me - that's the point.
All of it is free. No paywalls, no premium tier, no course to buy - the same material front offices read about, explained in the open.
Any single number that promises to end an argument. EPA, DVOA, and win probability are estimates built on models, and the explainers here spend as much time on where each metric breaks as on what it measures. I'm equally suspicious of tiny samples dressed up as trends - "this team is 7-2 in primetime" is trivia, not analysis - and of turnover luck and "clutch" narratives that are mostly regression waiting to happen.
The same honesty applies to the betting material: it teaches the math behind odds, expected value, and bankroll management - it does not sell picks, and no free website beats the sportsbook's vig for you. Read the disclaimer before putting money on anything, and if a number anywhere on this site looks wrong, tell me which page - every figure has a data pull behind it, so corrections are quick to verify and I fix them and say so.
I'm C. B. Zakarian. I write a family of data sites — sports analytics on one side, Roblox's trading economies on the other — all built the same way: public data, open methods, real charts, no invented numbers. The range isn't as odd as it looks; a Pythagorean win expectation and a virtual pet's trade value are the same problem in different clothes — noisy public numbers that reward careful measurement. Every site carries my name because I'd rather stand behind the work than hide behind a brand.
None of this would be possible without the open-source community that made NFL data public in the first place.
The foundation of modern NFL analytics. Ben Baldwin and Sebastian Carl's R package provides the data that powers the tutorials here.
The complete NFL data ecosystem including rosters, schedules, draft data, and more. A community-driven project.
Cooper's Python package brings nflverse data to Python users, expanding accessibility to another programming community.
The countless analysts on Twitter, Discord, and elsewhere who share knowledge and push the field forward.
Begin your NFL analytics journey with Chapter 1 - no prior experience required.