One football universe. One fan-intelligence layer. One argument.
CFB Zeitgeist is a college-football reference product. It combines a power+resume
ranking model that covers every NCAA team, FBS through Division III
with a layer that listens to what fans are actually saying about each program — belief,
respect, rivalry heat. The point is to support better arguments, not faster ones.
What you'll find
Power Rankings — a single board ranking every team across every level, with a Power side (predictive) and a Resume side (earned).
Heisman Tracker — a full-board Heisman model, not just a top-three list. Nowcast, Forecast, Win, Finalist, and Ballot probabilities for every real contender.
Team Pages — one per program, including a deep-treatment design for 158 profiled teams (mood, chronicle, savant, rivalry).
Player Pages — 17,000+ player profiles with percentile context and signature stories.
Two models run in parallel. The predictive power model asks how strong a team is at any given moment. The resume model asks what that team has earned so far — who they played, what happened, how the results compare to expectation. Keeping them separate matches one of the most useful distinctions in serious college-football analytics.
On top of the models, a fan-intelligence layer reads Reddit, mainstream coverage, betting markets, and recruiting signal. The result is what each fanbase actually believes, not just what the box score says.
Fans who want to argue with more receipts. Analysts who want one place to check both the model and the mood. Anyone who likes their college football served with a perspective.
About this site
CFB Zeitgeist is a personal hobby project — one person, public data sources, and a lot of Python.
It is not affiliated with the NCAA, any conference, any school, or any media company.
It is not professional journalism.
Some content on this site — specifically the team chronicle narratives and player story cards —
is generated with the help of local language models (AI). That content is labeled
"AI-generated · may contain errors" wherever it appears.
Language models can hallucinate: they can state things confidently that are false.
If you see something that looks wrong, it probably is — don't repeat it without checking a real source.
All analysis (rankings, model projections, fan-sentiment reads) is editorial opinion,
not professional sports reporting or financial advice.
Data sources are listed at /data-sources/.
