NFL · Data Science · Validated against PFF

The NFL film room,
rebuilt on data you can trust.

Player-tracking metrics validated against the industry standard. Coordinator schemes, matchup gameplans, career arcs, live camp buzz — every number carries its receipts, every limitation stated plainly. Calibration & scouting — not betting picks.

NFL Big Data Bowl tracking nflverse play-by-play 32 teams · 2016–2026 Honest disconfirmation
Why it's different

Data science, not just stats.

Most NFL stat sites show you box scores. This is built on player-tracking data and a discipline: every metric is benchmarked, and nothing is asserted that wasn't tested.

01

Real tracking data

Built on NFL Big Data Bowl full-22 player tracking + ball-flight — not box scores. We see who rushed, who covered, how open a receiver got.

5.4M+ frames
02

Validated, not asserted

Every tracking metric is benchmarked against PFF's charted equivalent. If our number doesn't track the standard, it doesn't ship.

+0.64 to +0.99 vs PFF
03

Scheme intelligence

Coordinator blitz & coverage DNA, formation gameplans, the matchup formation war — the scheme layer box scores can't see.

64 coordinators mapped
04

Honest by default

When a hypothesis fails a backtest, we say so plainly. The betting-edge question is closed — this is scouting and calibration, not picks.

Truth > a good story
Inside the film room

Six ways to read the game.

Every surface is live, interactive, and sourced — straight from the tracking data and the cube.

Research & Publications

Quantitative Football Whitepapers

High-frequency tracking telemetry, Bayesian kinematic filtering, and 2D spatial EPA surface modeling.

Browse All Technical Articles & Archive →
We publish the metrics that failed, too.

Separation didn't predict route grade. Roster value-weighting didn't beat turnover-only. The in-sample betting "edge" was selection bias. We tested them, they didn't hold, and we say so — because for scouting and calibration, an honest no is worth more than a confident maybe.