WR2026-05-30

The "high-aDOT sleeper" is a mirage. Downfield role doesn't move next year's points.

What the consensus says

Air yards are having a moment again. FantasyPros' wide receiver breakout list leans on average depth of target as the tell, flagging Alec Pierce as a buy on the second-highest aDOT in the league. Their "WR3s with WR1 potential" piece sells the same downfield-role story. FTN and FantasyLife both run dedicated air-yards trackers, and FantasyLife's explainer literally tells you "why you should pay attention." NBC ranks receiver-friendly QBs by how far they throw it, and PFF grades the deep-ball prospects up. Six-plus independent voices, one reflex: a receiver who runs deep and racks up air yards is a coiled breakout, undervalued because the catches have not shown up yet. Buy the aDOT, wait for the spike.

Falsifiable. So I beat on it.

The claim, in plain English

def predicate(row) -> bool:
    """In-cohort = deep-threat / downfield role: Y0 aDOT >= 13.0."""
    return (row["receiving_air_yards"] / row["targets"]) >= 13.0

A receiver is "in cohort" if his average target traveled at least 13 yards in the air, roughly the top quartile of qualified WRs. That is the downfield role the takesmiths are buying. The control is the whole game: a deep aDOT means nothing if it just rides along with volume, so I band every receiver by target share and ask whether the downfield guy beats the short-area guy at the same share of the offense. Conveniently, aDOT and target share barely correlate in this sample (negative 0.02), so the band split is clean.

How I beat on it

scripts/h45_deep_threat_wr_adot.py builds Y to Y+1 pairs from player_stats_season.csv (REG only, WR, 2015-2024 Y0, Y+1 reaching 2025, at least 6 games and 40 targets so the aDOT is a stable role measure, not a four-throw fluke). Inside each Y0 target-share band I compare next-year ppg for the deep-threat receivers against everyone else. Scored under half-PPR. Full framework: Session 97 two-gate spec. n = 681 pairs, a quarter of them deep threats.

What the data actually said

First swing, straight up: the deep threats averaged 0.47 ppg less in Y+1 than everyone else. Negative. The opposite of the breakout pitch, before any controls even kick in.

Because aggregates lie, I split by target share to see where, if anywhere, the downfield role pays:

Y0 target share	n_in	in ppg	n_out	out ppg	delta
0.25+ alpha	18	11.87	119	12.33	-0.46
0.20-0.25	37	10.61	153	10.52	+0.09
0.15-0.20	58	8.90	139	8.65	+0.26
0.12-0.15	37	7.44	62	6.66	+0.77
0.08-0.12	14	7.67	40	6.62	+1.04
under 0.08	1	5.77	3	4.29	n/a

Look at the shape. Among the alpha receivers, the ones who actually command an offense, the deep role is a slight negative: minus 0.46 a game. The faint positive only shows up at the bottom, in the 8-to-15 percent target-share tiers, the boom-bust dart-throwers, and even there it tops out at about a point a game and never clears the 1.5-ppg noise floor with a real sample on both sides. And the thinnest, lowest-volume cells are exactly where one big touchdown week swings a whole season's per-game line. That is not a signal. That is variance with a costume on.

I also checked whether 13.0 was a cherry. It is not. Re-run the cutoff at 11.5, 12, 13, or 14 air yards and the aggregate is negative every single time. Push it to a true 14-plus deep-ball specialist and the alpha band gets worse, minus 1.14. The deep-threat breakout does not survive contact with ten years of data at any threshold I tried.

What the engine already figured out

SignalTuned would not have made this bet, and here is the trap fooling the takesmiths. The engine reads aDOT, but only at the other end. Its slot_receiver archetype fires when a receiver's aDOT sits at or below 8.0, and it rewards that short-area possession role with a gentler post-peak decline, because space-creation and route precision age well. There is no high-aDOT archetype, no downfield-role forward bonus, nothing that turns "throws it deep" into "expect a jump." The engine treats aDOT as a description of how a receiver is used, not a prediction of whether he is about to pop.

The cohort validated that silence, and here is the mechanism the breakout pitch quietly skips. Air yards measure where the ball goes, not how often it gets there. A deep threat can rack up gaudy air-yards totals on three incompletions for every contested catch. Volume, meaning target share, is what actually fills a stat line. The breakout pitch swaps "had a lot of air yards" for "is about to earn a lot of targets." Those are different sentences, and the ten-year cohort does not confuse them.

What to do about it

Do not pay a breakout tax for a deep aDOT. A downfield role is a style, not an edge, and at the top of the depth chart it is a faint negative. The thing that actually predicts the WR leap is the boring one: target share, the offense feeding him more. The engine already prices that and ignores the air-yards noise. If your group chat is buying a deep burner because the aDOT is gaudy and the catches "have to come," the ten-year cohort says they mostly do not, and next year's per-game line is a coin flip. No engine change ships from this one. The engine was already holding the right hand.

Receipts

← Back to all posts