TE2026-05-24

Year 3 doesn't flip a TE switch. It nudges one.

What the consensus says

Pull up any 2026 dynasty TE column right now and you'll find the same arc. Year 1 is overwhelm. Year 2 is acclimation. Year 3 is when the lights come on. Bear Goggles On framed Colston Loveland's rookie finish as a unique launch point for 2026. RotoBaller's post-draft dynasty TE rankings lean on the same framing for Loveland, AJ Barner, and Isaiah Likely. Footballguys' dynasty TE tiers, FantasyPros' four-TE breakout list, PlayerProfiler's top-10 dynasty breakouts, and the Fantasy Footballers' TE lifecycle piece all run some version of it. Six independent voices, one calendar-shaped claim. So I tested it.

The claim, in plain English

The hypothesis: a TE whose Y0 was his 2nd NFL season produces higher Y+1 (year-three) ppg than a TE whose Y0 was a later career year, holding Y0 target share constant. The predicate is one line:

def predicate(row):
    return (row["season"] - first_nfl_season[row["player_id"]]) == 1

The "at the same target share" part is the whole game. Year-2 TEs tend to sit at slightly different depth-chart spots than journeymen vets. Skip the volume control and you measure depth-chart luck dressed up as a development arc.

How I beat on it

Every TE season 2015 through 2024 with at least four games and 15 targets, paired against the same player's next season, scored under TE-premium half-PPR (1.0 per reception, the TE bonus baked in). 482 pairs survived. 67 were year-two NFL seasons (the cohort), 415 were later vets. Six target-share bands. Within-band Y+1 ppg averages. Full framework at docs/RESEARCH_SESSION97_AUTONOMOUS_2026-05-13.md.

What the data actually said

First swing, aggregate: year-two TEs averaged 7.51 Y+1 ppg, vets 7.42. A delta of nine hundredths of a point. The "year 3 light flips on" claim wants something big and aggregate-shaped. It is not there.

Because aggregates lie. Year-two TEs cluster differently in target share than vets do, so the within-band view is the only one that means anything:

Y0 target_share	year-2 n	year-2 Y+1 ppg	vet n	vet Y+1 ppg	delta
18%+ alpha	9	11.6	94	11.0	+0.7
14-18%	13	8.9	66	8.2	+0.7
11-14%	10	8.0	78	6.6	+1.4
8-11%	17	5.4	71	5.8	-0.5
5-8%	15	5.1	95	4.8	+0.3
under 5%	3	7.8	11	4.1	+3.7

Direction is mostly positive. Mostly. Four of six bands tilt year-two's way, but the magnitudes are noise-floor stuff. The 11-14% band gets closest at +1.4 ppg, and even that just misses the pre-registered 1.5 ppg threshold with only ten year-twos against seventy-eight vets. The under-5% band shows +3.7 ppg but with three players in the cohort, which is the kind of small-cell number you do not bring to a draft. The 8-11% band actively flips negative.

Pre-registered Gate 1 bar: 1.5 ppg in at least one band with ten per side. We don't get there. The honest read is "a hair positive on average, no band cleanly above noise."

Here's the mechanism the calendar-shaped narrative papers over. A TE who breaks out in year three isn't breaking out because the calendar flipped. He's breaking out because a new coordinator got hired, or a veteran TE ahead of him got cut, or his quarterback finally started throwing to him on third down. The year-three label is a proxy for those events, and a coarse one. Strip the target-share control away and the "development arc" is mostly just a different distribution of opportunity in each group. The calendar is not the cause. The cause is the change in usage, and you already knew to look for that before you ran any of these numbers.

What the engine already figured out

The engine wouldn't have made this bet either. There's a block called rookie_development_ramp that applies a four-year multiplier to TE LTV on top of the age curve: 0.80 in year 1, 0.88 in year 2, 0.95 in year 3, then 1.00 from year 4 onward. Translated: the engine already assumes a year-2-to-year-3 lift of about eight percent of full TE LTV. Small, positive, calendar-aware.

Both pictures land in the same place. The cohort test says the population-wide year-2-to-year-3 lift exists but is small and noisy. The engine encodes a small, positive multiplier in the same direction. The "year 3 flips the switch" framing is the takesmiths' version of "the engine's existing seven percent bump." The narrative inflated a real-but-modest effect into a discrete event.

What to do about it

For the 2026 draft: don't bid on a TE because he's entering year three. Bid on a TE because something concrete about his opportunity is about to change. New OC who actually targets the position. A vacated TE1 ahead of him on the depth chart. A QB upgrade that lifts every pass-catcher's targets. Those are bets with real magnitudes attached. "He's a year-three guy" by itself is taking the calendar to a math fight.

Loveland's 2026 case isn't his year-three label. It's the 28.5% target share over Chicago's final four games last year, plus Ben Johnson's history of feeding the position. Barner's case isn't year-three either. It's whether Seattle keeps re-routing snaps through him after the offseason. Likely's isn't year-three. It's whether Theo Johnson actually loses the starting job.

Engine action: none. The rookie_development_ramp.TE block is already pointed in the right direction at the right magnitude. No diff this cycle.

So next time someone tells you a TE is "set up to break out in year three," ask what specifically changed about his target share. If they can't name the change, the label is doing none of the work.

Receipts

← Back to all posts