One registered speed feature and opportunity test. Fit on 2023 rookies; choose settings with player folds that also remove validation games from training. Freeze before evaluating 2024 rookies. Equal weight per receiver.

The conventional-test model gets the same time and route interactions. Repeated attempts occur in one session. Thousands of routes still represent only 45 receivers.

Actual throw timing, route and coverage are historical observations. These models cannot tell a coach before the snap to wait longer, change a route or substitute a faster player.

NFL peak speed uses only snap-through-throw tracking. Coordinate speeds use a seven-frame filter and omit three edge frames. The displayed drill is one attempt; the study feature is the median across attempts.

Both panels show measured receiver coordinates, with the same yard scale and starting time. Up is downfield. Sprint clips start when their recorded tracking starts; NFL clips start at the snap. These are not equivalent athletic starting conditions.

The receiver’s last observed position holds after a recording ends. NFL tracking stops at the throw. No opponent, quarterback or ball is invented.

Earlier and later are different real plays by the same receiver with matching route, coverage and lineup labels. Cushion and other circumstances can differ. They are illustrations, not controlled experiments or simulated alternatives.

Selection used 2023 training context and Combine speed only—not separation, NFL speed, target status, player fame or model success.