A quant second-opinion.
Draftanomics is built for the quant-curious NFL draft fan — the audience that already follows Brugler, Baldwin, Menon, and Zierlein, but wants the receipts in one place.
I'm Christopher Gardella, and I built Draftanomics to do one thing well: be the most-credible quant counterpoint to the consensus draft boards. Not a replacement for Brugler or Kiper — a complement, with two-decimal honesty and a public, out-of-sample track record.
What this site is.
A platform with three load-bearing parts: an out-of-sample backtest showing the model's board predicts real NFL career value better than the analyst consensus — the best a public-information model has done, a step shy of NFL war rooms; per-prospect tier predictions with bust probability calibrated and disclosed; and a comp engine that finds the NFL veteran whose pre-draft measurables and production most resemble a prospect's — statistical matches, not narrative ones.
What it isn't.
Not a mock-draft factory. Not a replacement for tape. Not a hype machine. When the model fails — and it does, in spots — Draftanomics ships the failure page and flags, right on the prospect's page, any position where the bust model isn't reliable enough to stand on its own. Those positions lean on the all-positions model instead — and the site tells you when that's happening.
How to read it.
- Every per-prospect bust probability ships with a plain label for how accurate that position's model has been historically. Trust the strong positions; apply your own judgment where the site says the model is on thinner ice. The full accounting is on the methodology page.
- Track-record is per-class. A 60% top-32 overlap in 2017 doesn't mean the model is right about 2026 — it means the methodology has held over time.
- The editorial is where I show the model the work — dissents, accuracy check-ins, and methodology writeups. Subscribe via RSS or email if you want it in your reader.
The wedge.
Draftanomics won't out-volume PFF, won't out-prose Brugler, won't out-reach ESPN. It can be the one place where the model is honest, the methodology is documented, and the receipts are public. That's the bet.