The five archetypes
Archetypes describe how a receiver is used, not how good he is. Average production is similar across all five; within each one, the receivers who get open produce the most. That is why every label comes paired with a separation tier.
How it works
Each row is one receiver-season with 200 or more routes (100 or more for 2026, which is in progress).
- Twenty-three stats go in: separation score and win rate against man and zone, aDOT, average depth of route, slot rate, YAC per reception, deep and contested target shares, designed-target rate, first-read rate, and the share of routes run on eleven route types.
- Every stat is compared with that season's receivers only, so a 2022 score and a 2025 score mean the same thing.
- Principal component analysis with a varimax rotation condenses the stats into six axes. Archetypes come from a Gaussian mixture model on the four role axes (Depth, Manufactured touches, Quick game, In-breaker). Separation is a modifier: plus separator at +0.75 or higher, limited at −0.75 or lower.
- Fit is the model's probability for the assigned archetype. Confidence combines fit with a bootstrap test: High means at least 70% fit and the same label in at least 70% of 100 refits on resampled players.
- Axis scores are standard deviations from that season's average. The bar shows direction, not quality; only Separation measures quality.
- Career archetype is the role a receiver held for the largest share of his 2022–25 routes. Only 55% of receivers keep the same archetype from one season to the next, so check the season history.
- Route mix reflects the offense as well as the player, and the five archetypes overlap. Read a label as "mostly this," not as a hard box.