Don't Look at My Team

TreVeyon Henderson — RB, NE (2026)

Verdict

FADE at ADP 58.7 / RB24 (FFC PPR mocks, 2026-07-07), medium confidence. The market's case is fair: a 23-year-old year-2 R2 pick who went 180-911-9 (5.06 YPC, +0.85 RYOE/att) as a rookie on the best offense in football, showed three RB1-overall-level weeks (24.3 PPG, W9–11) when Stevenson sat, and plays for a staff expected to grow his snap share past 50%. But the box score the price is built on was rented: 4 of his 9 rush TDs came from 52+ yards out (xTD ≈ 6.5 vs 9 actual), and in the 14 games both backs were healthy his role was 37.3% of snaps, 20.6% of third downs, 24.4% of two-minute dropbacks, and 6 of 23 inside-5 carries — then it *shrank* down the stretch (38.8% snaps, 14.6% third downs W15–18) and cratered in the playoffs (41/38/6/39% snaps; 2.5 YPC). Why the market is wrong: pick 59 pays the RB21 season plus year-2 growth, but the year-2-leap screen fails on its own terms (late-season snap share falling, not ≥60%; competition not departing — Stevenson is contractually locked for 2026, dead cap > savings), the passing-down path is gated by a 23.6 PFF pass-block grade (39th of 43), and the June 2026 beat still frames it "1A/1B" with Stevenson the "lead back... ground-and-pound guy" — coach-speak the methodology says to price as *committee* (rb.md §7, §12). Fine player, wrong price: he's a happy pick a round-plus later (~pick 70+, next to Stevenson's own 72.0 ADP).

Bull case

The talent is real and line-independent: +0.85 RYOE/att (elite band), 5.6% breakaway rate, 5.06 YPC, 9 rush TDs at age 23 with R2 capital — and the W9–11 audition (82.5% snaps, 89% of dropbacks, 24.3 PPG) proved the ceiling is RB1-overall weeks, not hypothetical. Stevenson has missed games in both of the last two seasons; the contingent case activates regularly.
Best possible environment: the league's best offense by EPA/play, a 9.5–10 win total, gap/duo scheme fit, and a leading-by-7+ snap share of 54.6% — his carry volume is script-protected, and every TD-equity input (red-zone trips, light boxes) is elite. A modest role bump on this offense outkicks a bigger role on a bad one.
The receiving gate is coachable, and he knows it: 0.154 targets per on-field dropback (above Stevenson's 0.109) says he earns work in the pattern; pass-pro (23.6 PFF) is the one skill keeping him off passing downs, he named it as his offseason focus (June 2026), and Stevenson turns 28 with no guarantees beyond this season. If camp reports show him taking two-minute reps, the whole profile re-rates.

Bear case

The role the price assumes never existed when both backs were healthy: 37.3% of snaps, 20.6% of third downs, 24.4% of two-minute work, 6 of 23 inside-5 carries, 33.3% of one-score snaps — and it *fell* late (38.8% snaps, 14.6% third downs W15–18) before bottoming out in the playoffs (6% snap share in the AFC Championship; Stevenson 94%). The staff's revealed preference in the biggest games of a Super Bowl run was overwhelming, and the incumbent it preferred is locked for 2026 with dead cap exceeding savings.
~20 points of the 2025 line were variance: 4 TDs from 52+ yards (xTD ≈ 6.5 vs 9 actual), a 41.7% success rate, and a 3-game bellcow window supplied the RB21 finish the ADP is anchored to; strip them and the both-active pace (13.7 weighted opps/g, 3.07 HVT/g, 7 of 17 games under 8 PPR) is a boom-bust flex priced two rounds ahead of his backfield partner — who out-graded him as a rusher (+1.36 RYOE/att), receiver (9.3 vs 5.3 y/tgt), and pass-blocker (67.6 vs 23.6).
PPR pays targets, and his passing-down path is gated: 23.6 PFF pass-block grade (39/43), 2.1 targets/g both-active at −0.02 air yards, 8 third-down targets all season, and the checkdown tree just got A.J. Brown added to it. Per rb.md §9 the receiving projection stays capped until pads-on evidence says otherwise — a back with ~2 targets/g has a floor of zero in losses, and his 5-games-under-6.0-PPR 2025 already showed it.

Projection & comps

Bottom-up (team inputs from data/team-profiles/NE.md, built 2026-07-07: ~62 plays/g, ~60% dropback rate, ~25 designed rush att/g ≈ ~23 RB carries/g, win total 9.5–10, positive script):

Scenario	Games	Carries	Rush yds	Targets → rec/yds	Total TD	PPR pts
Floor (p20)	15	~150 (44% RB share, role static)	~630 (4.2 ypc)	34 → 26 / 160	5	~135
Median (p50)	16	~172 (~47% share, modest growth)	~775 (4.5 ypc)	44 → 35 / 245	7	~185
Ceiling (p80)	16	~210 (Stevenson misses 4+ or share flips)	~985 (4.7 ypc)	56 → 45 / 310	10	~245

TDs anchored to xTD, not 2025 actuals: 2025 rush xTD ≈ 6.5 (8 inside-5 carries × ~0.42 + 13 carries from the 6–10 × ~0.12 + ~1.5 field-expectation on 180 carries for an elite-breakaway profile; pbp-derived, pulled 2026-07-07) vs 9 actual — he banked ~2.5–3 TDs of long-run variance (TD yardlines: 2, 5, 5, 7, 7, 52, 55, 65, 69). Median 7 assumes similar usage on an elite offense with normal long-TD luck; Stevenson's inside-5 edge (12 of 23 both-active) caps the goal-line term.
Floor = W15–18/playoff role persists + TD regression; ceiling = the W9–11 audition recurs for a month-plus (Stevenson has missed games in each of the last two seasons: 15 of 17 in 2024, 14 of 17 in 2025 — weekly.csv 2024/2025) or Henderson's camp pass-pro/route work wins him passing downs.
Games-played risk: medium — RB baseline; he himself played all 21 games (17 REG + 4 POST) as a rookie; W17 concussion and AFCC-week shoulder both closed as full practice participation (injuries.csv, 2025).
Comps (role/range sanity check, approximate PPR): Jahmyr Gibbs 2023 (182-945-10 + 52 rec, ≈240 — the ceiling shape: explosive 1B who forces the flip); Tony Pollard 2021 (130-719-2 + 39 rec, ≈170 — explosive 1B stuck behind a contract incumbent); A.J. Dillon 2021 (187-803-5 + 34 rec, ≈187 — committee on a top offense); Jaylen Warren 2023 (149-784-4 + 61 rec, ≈180); Zach Charbonnet 2024 (135-569-8 + 42 rec, ≈176 — value arrives only in the starter's absences).
External projections: data/projections/ does not exist. Fantasy Life (fetched 2026-07-07) projects 163 rushes / 43 targets, ranks him RB26 vs an RB21–24 market price and calls the takeover "unlikely"; DraftSharks NE preview (fetched 2026-07-07) projects a 39.5% vs 35.5% carry-share edge and frames him "low-end RB2 with top-10 upside if Stevenson misses time." My median (185, ~RB24–26) agrees with both — the disagreement is with the price, not the projections.

Usage profile — opportunity core (2025 rookie season, 17 REG games)

Cache numbers: nflverse via nflreadpy, pulled 2026-07-07; play-level splits derived 2026-07-07 from 2025 pbp (nflreadpy) joined to participation.csv on game_id+play_id (1,046/1,046 NE rush+dropback plays joined). "Both-active" = the 14 games Stevenson also played (W1–8, 12–13, 15–18; Stevenson missed W9–11, toe).

Metric	Value	Band	Read
Snap share	45.9% season (snap_counts.csv) — but 37.3% both-active vs Stevenson 59.3%; W9–11 (Ste out) 82.5%	Concern (both-active)	The season number is a blend of two different players: a 37% committee back and a 3-week bellcow
Opportunity share	48.7% season (222 of 456 backfield opps); both-active carries dead even (133 vs 131)	Good (barely)	Carries were a true 50/50; everything else tilted Stevenson
Weighted opps /g (c + 2.5t)	16.8 season; ≈13.7 both-active (133 c + 2.5×30 t / 14 g)	Concern–Good line	Below the 18 "Good" floor in the honest split
High-value touches /g	3.71 season (2.47 tgt + 1.24 inside-10 car); 3.07 both-active	Below Good (4–6)	The scoring engine ran on explosives, not HVT
Inside-5 carry share (team)	28.6% season (8/28); both-active 6 vs Stevenson 12 (of 23); inside-10 both-active 13 vs 17 (of 39)	Concern	Goal line belongs to Stevenson; Henderson's TD access is the home run
Third-down snap share	34.8% season; 20.6% both-active (Ste 83.1%); two-minute dropbacks 24.4% both-active (Ste 74.4%); 8 third-down targets all season	Concern	He leaves the field on exactly the downs that carry PPR value
Routes /g · route participation	Route counts UNVERIFIED in cache; proxy: on-field for 34.4% of dropbacks both-active (44.5% season). DraftSharks charting (fetched 2026-07-07): route rate 29.2% both-active vs Stevenson 48.8%	Concern	2×2 read: modest opp share and low snap share on passing downs — the fragile quadrant
xFP	Provider xFP UNVERIFIED. Actual 12.13 PPG (206.2, RB21 total / RB25 PPG min-8-games) with TD output ~2.5–3 above xTD	~RB2/3 border	Usage-based expectation ran below the actuals — negative hidden value

Receiving profile (§3): 42 targets in 17 g (2.5/g season; 2.1/g both-active — Concern side of the 3–5 band). 35-221-1, 5.3 yds/target (31st of 43 qualified RBs), 47.6% receiving success (18th), passer rating when targeted 96.5 (22nd) — vs Stevenson's 9.3 y/tgt (1st), 59.5% (1st), 123.5 (2nd) (Fantasy Life charting table, fetched 2026-07-07). Target mix from participation.csv route charting: swings 12, hitch/curls 9, quick outs 9, screens 8 — designed-manufactured underneath work at −0.02 average air yards; only 4 of 42 targets came trailing by 7+. One genuine bull nugget: targets per on-field dropback 0.154 vs Stevenson's 0.109 (pbp join, 2026-07-07) — when he's actually in the pattern he earns work; the problem is the field time, and that's gated by pass protection: PFF pass-block grade 23.6, 39th of 43 RBs (Fantasy Life, fetched 2026-07-07) — squarely under the <55 gate in rb.md §9 that says cap the receiving-role projection until proven otherwise. Henderson said it himself at June 2026 OTAs: "The biggest thing I want to continue to develop, blocking and... running routes out of the backfield, catching the ball" (SI/On SI, 2026-06-24).

Game script (§4): Season snap-play shares by score state (pbp × participation join, 2026-07-07): leading 7+ 54.6% · one-score 39.4% · trailing 7+ 43.8%; Stevenson flips it (one-score 56.8%). Henderson is the *clock-kill and blowout* back on a team that led by 7+ on 35.2% of its snaps — the 9.5–10 win total (DraftKings via team profile, 2026-06) protects his carry volume, so the script prior is genuinely supportive. But the one-score number is the tell: in competitive game states — most of every game — the staff trusted Stevenson. Weekly volatility follows: 7 of 17 games under 8.0 PPR points (5 under 6.0) against three games of 28+. That's a boom-bust flex, not a weekly starter, until the role changes.

Efficiency (§5) — separating the back from the line:

Metric	Value	Band	Source (as-of)
RYOE /att (NGS)	+0.85 (+149.0 total on 180 att)	Elite (≥+0.7)	ngs_rushing.csv, pulled 2026-07-07
Breakaway rate (15+ yd)	5.6% (10/180)	Good, near-elite	derived from 2025 pbp, 2026-07-07
Rush success (EPA>0)	41.7%	Below avg	derived from 2025 pbp, 2026-07-07 (Stevenson 37.4% behind the same line)
MTF (rushing)	30 (0.167/att; receiving MTF UNVERIFIED)	Good (low end)	PFF via web search, fetched 2026-07-07
YAC /att	3.5	Good–Elite	PFF via web search, fetched 2026-07-07 — treat with caution: DraftSharks separately ranked him 1st of 49 in yards before contact/att, and both can't sit high at 5.06 YPC
8+ in box	30.6% of att	context	ngs_rushing.csv — Maye's passing game bought decent boxes
PFF grades	rushing 75.0 (28/55), overall 68.9 (39/55)	mid	PFF via web search, fetched 2026-07-07

Read: the RYOE and breakaway numbers say the explosiveness is real and line-independent — but it's a *spike-shaped* profile (boom runs, 41.7% success rate), NGS rush%-over-expected just 39.8%, and per the evidence hierarchy a rookie-year efficiency level needs a second season before it's believed. Meanwhile Stevenson posted +1.36 RYOE/att (ngs_rushing.csv) behind the same line — there is no performance case for benching the incumbent, which is exactly what the takeover thesis quietly requires. Note the playoff sample: 2.5 YPC across 4 games (Athlon, fetched 2026-07-07).

Context (from `data/team-profiles/NE.md`, built 2026-07-07)

Elite, stable offense: McDaniels year 2 (calls plays), Maye entrenched (MVP runner-up), +0.157 EPA/play best-in-class 2025, win total 9.5–10, ~25 designed rush att/g. Positive scripts feed the run game — the environment is the best part of this profile.
OL: mid-tier, upgraded (PBWR 13th / RBWR 12th 2025; Vera-Tucker added at LG, Campbell back from knee at LT with a camp flag). No YPC headwind, modest tailwind.
Scheme: gap/duo-heavy with inside-zone mix, heavy 2-back/FB usage; McDaniels "tailors runs to the back" — both backs fit; no zone/gap mismatch flag.
Committee (§7 math): contested 1A/1B, and the 2025 ledger says Stevenson was the 1A where it counts — note this eval's play-level derivation *disagrees with the team profile's "passing downs: Henderson" line*; the joined pbp/participation data (83.1% vs 20.6% third-down both-active, 74.4% vs 24.4% two-minute) is authoritative per the evidence hierarchy. Stevenson is under contract through 2028 with no 2026 out ($7.42M cap, $8.05M dead — Spotrac/OTC via web, fetched 2026-07-07), age 28 (born 1998-02-23, rosters.csv), coming off his most efficient season. No RB capital added (R7-245 Jam Miller only). June 2026 beat: "Stevenson and Henderson... 1A and 1B" (patriots.com camp preview / Yahoo minicamp primer, June 2026); Stevenson the "lead back," Henderson the change-of-pace/home-run threat expected to pass 50% of snaps (SI, 2026-06-24 — analyst expectation, not a team statement). 2×2 placement: mid standalone / very high contingent — the contingent half is real league-winner material (24.3 PPG audition), but at pick 59 you're paying for it up front. Ambiguous-backfield check: Henderson 58.7 + Stevenson 72.0 both drafted at meaningful capital out of one backfield whose 2025 total was ~426 RB PPR points; Henderson's price implies ~48–50% of that pie *plus* growth — i.e., the market has pre-resolved the ambiguity in his favor while every recency signal (W15–18, playoffs, beat framing, contract) points the other way.
Target competition rose: A.J. Brown (trade, 2026-06-01) + Doubs ($39M gtd) + Henry's 18.1% share — RB target share was 17.2% in 2025 and the checkdown tree now competes with a true alpha. Not a growth vector for a 2-target/g back.
Pedigree/gates (§8–9): age 23 (24 on 2026-10-22; Sleeper file 2026-07-07), 2025 R2 P38 (rosters.csv) — capital buys him 2–3 years of opportunity, which is why AVOID is off the table. Odometer clean: 667 college touches over 4 years at Ohio State (590 car + 77 rec — Ohio State/Wikipedia via web, fetched 2026-07-07) + ~248 pro touches ≈ 915. 77 college receptions clear the three-down prior; the 23.6 pass-block grade is what's blocking it in the pros.
Archetype (§10): explosive committee back / change-of-pace with elite contingent value — priced here like an ascending lead back.

Tripwires (re-run on any)

Stevenson injury, trade, or release → verdict flips immediately (TARGET, likely MUST-HAVE at this price) — the W9–11 tape is the proof of concept.
Pads-on camp/preseason reports (late July–Aug 2026) put Henderson in the two-minute/third-down package or credibly report a pass-pro leap → the §9 gate lifts; re-run, likely HOLD→TARGET.
ADP drifts past ~70 (round 6, 12-team, near Stevenson's 72.0) → price meets profile; flips to TARGET on cost alone.
Preseason first-team usage shows Henderson leading early downs AND taking goal-line reps → the 1A framing was wrong; re-run.
Stevenson decline signals (age-28 burst complaints from the beat, camp snap reports flipping the "top dog" framing) → re-run toward HOLD/TARGET.

Sources

data/stats/2025/ — weekly.csv, rushing.csv, receiving.csv, snap_counts.csv, ngs_rushing.csv, injuries.csv, rosters.csv, participation.csv, pbp_summary.csv (nflverse via nflreadpy, pulled 2026-07-07; derived season tables REG-only). data/stats/2024/weekly.csv for Stevenson 2024 games.
Derived play-level splits (both-active committee shares, third-down/two-minute participation, score-state snap shares, inside-10/5 team shares, breakaway, EPA success, opportunity share, TPRR proxy, TD yardlines, target route mix) from 2025 pbp via nflreadpy joined to participation.csv on game_id+play_id, computed 2026-07-07; 1,046/1,046 NE rush+dropback plays joined; snap-share outputs reconcile with snap_counts.csv (45.5% vs 45.9%).
data/adp/adp_ffc_2026-07-07.csv — Henderson 58.7 / RB24; Stevenson 72.0; A.J. Brown 14.5; Maye 65.2 (FFC PPR mocks, 2026-07-07).
data/raw/sleeper/players_2026-07-07.json — age 23, DOB 2002-10-22, Ohio State, years_exp 1, depth_chart_order 1 (as-of 2026-07-07).
data/team-profiles/NE.md (built 2026-07-07) — McDaniels/Maye context, OL, scheme, win total 9.5–10 (DraftKings via CBS), plays/pass-rate projections, vacated/arrival math, RB committee read. Discrepancy noted: profile's "Passing downs: Henderson" line is contradicted by this eval's play-level derivation (Stevenson 83.1% of both-active third downs); the computed usage data governs.
Fantasy Life "Who Will Lead The Patriots Backfield" (fetched 2026-07-07): both-active touches 162/158, W10–18 Henderson 17.6 PPG (RB7), y/tgt 9.3 vs 5.3, receiving success 59.5% vs 47.6%, PFF pass-block 67.6 (5th) vs 23.6 (39th), FL projections (163 rush/43 tgt, RB25) and RB26 rank; takeover "unlikely."
DraftSharks NE preview (fetched 2026-07-07): both-active snap shares 59.3/37.5%, route rates 48.8/29.2%, inside-5 shares 50/27.3%, 2026 carry-share projection 39.5/35.5%, "low-end RB2 with top-10 upside if Stevenson misses time."
SI/On SI "Why TreVeyon Henderson Could Be Patriots' 2026 Breakout Star" (2026-06-24): >50% snap expectation, Henderson OTA quote on blocking/routes, Stevenson "lead back... ground-and-pound guy."
patriots.com Training Camp Preview: Running Backs + Yahoo minicamp primer (June 2026): "1A and 1B" framing, RB room (Jennings, Larison, Miller, Montgomery); minicamp June 9–11, 2026.
Athlon (fetched 2026-07-07): playoff 2.5 YPC.
Spotrac / Over The Cap via web (fetched 2026-07-07): Stevenson 4yr/$36M ext (2024), $17.1M gtd; 2026 cap $7.42M, dead cap $8.05M — no 2026 out.
PFF via web search (fetched 2026-07-07): Henderson overall 68.9 (39/55), rushing 75.0 (28/55), 30 MTF rushing, YAC/att 3.5 (flagged: conflicts with DraftSharks' yards-before-contact claim; treat with caution).
Ohio State Buckeyes site / Wikipedia via web (fetched 2026-07-07): college career 590 carries / 3,761 yds / 77 rec / 853 yds / 48 TD (2021–24).
League scoring: full PPR / 4pt pass TD / no TE premium assumed — methodology/league-settings.md placeholders unconfirmed.