The TE1-on-his-team premium is just target share looking in a mirror

What the consensus says

CBS Sports ranks Brock Bowers as TE1 in dynasty with a specific note: "he dominates his team's passing game" and "new coordinator Klint Kubiak could mean one of the highest first-read target shares in the league." Fantasy Life says "Bowers is the only show in town for this passing game." RotoBaller has him as TE1 in dynasty post-draft; Dynasty Nerds price him at the top of the position in startup ADP.

The argument running through all of it: Bowers and Trey McBride are worth the premium partly because they are their team's top receiver, full stop. McBride ran 694 routes in 2025, led every tight end in catches (126), yards (1,239), touchdowns (11), and targets (169). The Cardinals built around him. Bowers drew 153 targets on a Raiders offense that had nobody else to throw to. The claim is that this structural dominance of the target tree adds something on top of the volume.

Similar logic drives the Colston Loveland hype post-DJ-Moore trade. The Bears cleared their #1 receiver and Loveland stepped in. Owning the tree, in this framing, is the signal.

The claim, in plain English

If target tree ownership is a real premium signal, it should show up inside target share bands. At equal Y0 target share, a TE ranked top-3 in his team's overall target distribution (across all pass catchers, all positions) should produce more in Y+1 than a TE ranked fourth or lower.

def predicate(row: pd.Series) -> bool:
    """Cohort = TE ranks top-3 by total targets among all pass catchers
    on the same team in the same season. rank=1 is the team's most-targeted
    player regardless of position."""
    return bool(row["team_target_rank"] <= 3)

The test: does ordinal rank add anything beyond the share number already in the band? If Bowers' 24 percent share is already doing all the work, his rank-1 status should wash out once we control for it.

How I beat on it

scripts/h67_te_top2_team_target_pecking_order.py aggregated 2022-2024 weekly data to season level for all pass catchers, computed ordinal target rank within each team-season, then filtered to TEs with at least six games and 30 targets. Sixty-seven TE Y0->Y+1 pairs. Thirty-eight qualify as top-3 targets on their team. Twenty-nine do not. Framework: two-gate spec from Session 97. Half-PPR with TE reception premium (1.0 per reception), Dynasdeez 10-team format.

What the data actually said

First swing: aggregate. Top-3 TEs averaged 11.73 ppg in Y+1. Non-top-3 TEs averaged 9.82 ppg. That is a 1.91-ppg gap. The team-rank premium looks real.

Aggregate first. Bands second. Because aggregates lie.

The top line is the tell. Every TE with a ts>=18% season is a top-3 target on his team. There is no comparison group at high share because you cannot have high share without high rank. They are the same thing. Travis Kelce at 24.6% share, Sam LaPorta at 21.2% share, Evan Engram at 23.6% share -- all rank 1 or 2 on their respective teams. The "premium" is entirely that correlation.

In the one band where the groups actually overlap (14-18%), the direction reverses hard. Top-3 TEs at 14-18% share produced 8.69 ppg the next year. The one non-top-3 TE at the same share produced 12.70 ppg. That is a -4.01 gap in the opposite direction. Mark Andrews in 2022 (13.5 ppg, rank 1, ts 24.6%) and 2023 (13.5 ppg, rank 3, ts 12.8%) sits on both sides of this table across two seasons. The rank changed more than the quality.

The reason the direction reverses in the overlap band makes intuitive sense once you see it. A TE who ranks top-3 on his team at only 14-18% share is typically on a run-heavy or low-volume passing offense. He is the best receiver on a team that does not throw much. A TE outside the top 3 at the same 14-18% share is typically on a high-volume offense with a crowded receiver room. The offense passes more, the targets are real, the scheme is pass-first. Same share, very different ceiling.

Gate 1 verdict: fail. No band clears both the n>=10-per-side and the absolute-delta-1.5-ppg thresholds in the expected direction. The aggregate 1.91-ppg premium is a target share artifact, not a team-rank premium.

What the engine already figured out

The engine wouldn't have made this bet either. Here is the trap fooling the takesmiths.

The engine does compute team_target_rank for every player in the dynasty payload, and it surfaces that rank in trade analyzer comparisons. But it deliberately excludes the rank from the projection score. No config block routes team_target_rank into weighted_ppg or the age-curve blend. The engine looked at the same information available to every dynasty writer and decided it was already downstream of target share.

Our cohort test validates that call. Team rank is a consequence of target share, not a cause of Y+1 output. When you control for share, the rank contribution is zero to negative.

What to do about it

The concrete call: when you evaluate a TE in a trade or auction, the two numbers that matter are target share and offensive pass volume (team pass attempts per game). A TE at 18 percent share on a team throwing 38 times a game is worth more in dynasty than a TE at 18 percent share on a team throwing 26 times a game, regardless of what rank he holds on his respective roster.

This changes a few specific evaluations. Bowers' premium rides on how pass-heavy the Raiders offense becomes under new head coach Klint Kubiak, not on his rank as the team's top option. The premium is legitimate only if the pass volume supports it, not because Bowers is "the only show in town." George Kittle ranked 2nd or 3rd on the 49ers for most of 2022-2024 while producing elite ppg because the 49ers ran a pass-heavy scheme with high volume. His rank did not matter.

The team target rank is a label. The share and the offense's pass volume are the assets. Buy those.

← Back to all posts