Don't Look at My Team

RJ Harvey (RB, DEN) — 2026

Verdict

TARGET at 88.4 (FFC PPR mocks, 2026-07-07 — RB32, mid-8th round in 12-team). Harvey finished RB20 overall (206.6 PPR, 12.2/g) as a rookie in a 1B role and averaged 15.2 PPG on ~60% snaps in the 7 games after Dobbins' foot injury — and the market is now pricing him twelve RB slots below last year's finish. Why the market is wrong: it is reacting to three headlines — the Dobbins re-signing, offseason shoulder surgery, and inevitable TD regression — while ignoring that the role that produced RB20 (the Joker receiving job, ~3.4 targets/g, and near-half of the team's inside-5 carries) returns intact in a top-3-volume offense, and that the contingent scenario is nearly free: the 1A ahead of him is a 27-year-old who has never finished a season healthy since 2020 (15/0/8/1/13/10 games, 2020–25). At pick 88 you pay roughly for Harvey's standalone floor and get the Dobbins-outage upside — already demonstrated at a 15+ PPG clip — at no charge. Confidence is medium, not high, because his rookie rushing efficiency was genuinely bad (−0.64 RYOE/att, 36% success rate) and the shoulder cost him most of his install offseason under a first-time play-caller.

Bull case

The market is charging a flex price for a demonstrated 15.2-PPG lead role one Dobbins injury away — and Dobbins has missed time in six straight seasons. Contingency at this probability is usually the most mispriced asset in RB pricing, and Harvey's version comes with a standalone RB2/3 floor attached (12.2 PPG as a rookie 1B).
The scoring engine is touch quality, not touch count: 4.3 HVT/g full-season (5.9 in the lead stretch), 48% of team inside-5 carries, and a designed goal-line receiving role — in a 9.5-win, top-3-plays offense behind PFF's No. 1 OL. Usage this valuable at RB32 is exactly the "HVT under a committee label" green flag (rb.md §12).
The receiving trajectory is the year-2 signal: dropback participation climbed 32% → 57% across the season, targets held at 3.4/g in both phases, and the Joker role is play-caller-designed (Payton system retained). Rising late-season routes precede next-season target jumps (rb.md §11), and receiving volume is script-proof.

Bear case

He may simply not be good at the base job. −0.64 RYOE/att, 36% success rate, 2.7 YAC/att, 49th-of-55 PFF rushing grade — behind the best run-blocking environment in football, while his 27-year-old teammate posted +1.08 RYOE on the same line. If the runner doesn't improve, Denver has no football reason to ever hand him the 1A, and the ceiling scenario dies even if Dobbins gets hurt (a wk11–18 rerun capped at 60% snaps with Coleman/Badie eating the rest).
The 2025 points flattered the profile: 12 TDs against ~9–10 xTD (two rush TDs of 38–40 yards from a 2.7% breakaway profile), worth ~15–20 PPR points of regression, and he never owned obvious passing downs (16% of 3rd-and-5+ even as the lead) — so the "PPR back" label is scheme-dependent, not script-dependent, and a new play-caller could de-emphasize the Joker package.
The offseason went against him: shoulder surgery wiped out most of his OTA install reps in the very offseason a first-time play-caller rebuilt the offense, the team paid Dobbins $8M guaranteed to stay 1A, and spent a top-110 pick on another back. Three separate organizational votes of something short of confidence.

Projection & comps

Bottom-up, full PPR, 2026 team volume from data/team-profiles/DEN.md (~64.5 plays/g, ~25 rush att/g, ~35.5 pass att/g, ~20% RB target share → ~120 RB targets):

Scenario	Games	Carries	Rush yds	Tgt	Rec	Rec yds	TD	PPR
Floor (20th) — Dobbins plays 15+, Coleman eats passing-down work, shoulder lingers	14	~95	~370	~38	~30	~215	5	~120
Median (50th) — "similar roles to last year": 1B for ~11 games, lead for ~5 (Dobbins misses time, per his career base rate)	16	~150	~605	~54	~43	~350	8.5 (xTD-anchored)	~190
Ceiling (80th) — Dobbins down early or Harvey wins 1A; 10+ games in the wk11–18 role	17	~200	~840	~70	~56	~430	11	~255

TDs anchored to usage, not his 12 actuals: bucket-based xTD from his 2025 usage ≈ 9–10 (10 inside-5 carries ≈ 4.5 xTD, other carries ≈ 2, plus goal-line targets ≈ 2.5–3) vs 12 scored — a ~2–3 TD overshoot, internal estimate from pbp (nflverse, pulled 2026-07-07). Provider xFP: UNVERIFIED (no export in data/raw/).
Ceiling sanity check: his actual wk11–18 pace was 15.2 PPG × 17 = 259 — consistent with the 255.
Games risk: medium — labrum surgically repaired with "full go" expectation for camp (see §4), normal RB baseline, no workload flag (≈208 career pro touches).
Comps (role/profile): Tony Pollard 2021→22 (explosive 1B behind a brand-name starter, paid off when volume arrived), Zach Charbonnet 2024 (spike-week 1B behind a fragile lead), Chase Brown 2024 (day-2 back who took the job mid-season), Rachaad White 2023 (poor rushing efficiency, RB1-adjacent season on receiving volume + TDs), Alvin Kamara 2017 (the Payton Joker archetype — stylistic ceiling comp only; Kamara's efficiency was in another universe).
No external projections in data/projections/ (directory absent) — no sanity-check disagreement to report.

Usage profile

All 2025 REG, nflverse pulls dated 2026-07-07 (data/stats/2025/, raw pbp + participation join computed this run). Split shown because Dobbins' foot injury ended his season after Wk 10 — per rb.md §2, the wk11–18 new-role split is the real signal for the contingent scenario; the full season is the 1B baseline.

Metric	Full season (17 g)	Wk 1–10 (Dobbins active)	Wk 11–18 (Dobbins out, 7 g)	Read
Snap share	42%	29%	60%	Concern→good; never cleared 68% even as the lead
Opportunity share	43% (204/471)	~29%	64% (126/196)	Concern→good; McLaughlin+Badie still took ~35% with Dobbins out
Weighted opp/g (carries + 2.5×tgt)	17.1	~12	24.1	Below-good→near-elite
High-value touches/g (tgt + inside-10 carries)	4.3	~3.2	5.9	Good→near-elite; the scoring engine is real
Inside-10 / inside-5 carry share (team)	42% / 48%	18% / 27%	63% / 70%	Good→elite; he led DEN with 7 rush TD despite the committee
Third-down snap share	29%	26%	34%	Concern — and on 3rd-and-5+ only 16% (Badie: 50%)
Two-minute snap share	22%	12%	38%	Badie owned it (44–53%) all season
Dropback participation (route proxy)	43%	32%	57%	Good→elite trajectory; routes/g UNVERIFIED (no route-level data), proxy = on-field on team dropbacks
TPRR (proxy: tgt ÷ on-field dropbacks)	~0.20	—	~0.17	Good band; true TPRR slightly higher (proxy denominator includes pass-pro snaps)
xFP / PPG	12.2 PPG actual (RB20 total, RB24 PPG)	10.0 PPG	15.2 PPG	Actual ran ~1 TD/6wk hot vs usage; provider xFP UNVERIFIED

Receiving profile (§3): 58 targets (3.4/g), 47-356-5, 81% catch rate, aDOT 0.48 — effectively all of his receiving yardage came after the catch (360 YAC on 356 yards). This is a *designed* Joker role (screens, swings, goal-line angle routes — 3 of his 5 receiving TDs were targets at the opponent 1–5), not checkdown leakage: play-caller commitment, sticky while Payton's system stays. The caveat: obvious-passing-down work (3rd-and-long, two-minute) belonged to Badie all season, so the receiving role is scheme-fed rather than script-fed.

Efficiency (§5) — the bear section: 3.70 YPC, −0.64 RYOE/att (concern, <−0.3), 36.1% rush success rate (concern, <40%), −0.156 EPA/rush, breakaway rate 2.7% (4 runs of 15+; marginal), 8+-box rate 32.9% (ngs_rushing.csv + pbp). PFF: 68.3 overall, 69.2 rushing grade (49th of 55 qualifiers), 25 MTF as a rusher (~0.17/carry — decent), 2.7 yards after contact/att (below the 3.0 good line) (PFF via Mile High Report rookie-class review, 2025 season, searched 2026-07-07). The damning contrast: Dobbins on the same No. 4-RBWR line posted 5.05 YPC, +1.08 RYOE/att, 45.8% success, 7.8% breakaway. The line wasn't the problem. Per the evidence hierarchy this is one rookie season and usage outranks it — but it is exactly why Denver re-signed Dobbins, and it caps how hard we project a true 1A takeover on merit.

Game script (§4): He does *not* leave the field trailing — 34% participation when down 7+ vs 42% overall (60% vs 60% in the lead stretch). The Joker role is script-proof; DEN's 9.5 win total (DraftKings via CBS, 2026-07-01) and top-5 PROE prior make the environment a plus either way. Projection moves only modestly with team quality — this is a receiving-role back, not a positive-script grinder.

Committee 2×2 (§7): High standalone (locked Joker + goal-line share ≈ 10–12 PPG floor role) and high contingent (demonstrated 15.2 PPG when Dobbins went down). That's the good quadrant. Succession isn't perfectly clean — McLaughlin/Badie held ~35% of snaps even in the lead stretch and R4 Coleman joins the passing-down fight — which is what keeps the ceiling at 255 rather than 300. Ambiguous-backfield pricing check: backfield 2025 PPR total ≈ 405 (Harvey 206.6 + Dobbins 115.9 + satellites ~83, rushing/receiving.csv); Harvey's ADP-implied points at RB32 (~150–170) sit *below* his ~45–50% allocated share (~180–200) → the market is not overpaying for resolution; the trap doesn't fire. Dobbins at 95.9 is priced as a co-1A — only one of these prices is right, and the age-27 + injury-history + worse-receiving-role profile says it's Dobbins' that's too high.

Age/workload/gates (§8–9): Age 25 (b. 2001-02-04 — Sleeper players file, 2026-07-07; old for year 2, so the year-3+ prime window is shorter than typical), ~208 career pro touches (193 REG + ~15 POST) — no mileage flag; heavy college odometer (~600+ carries at UCF) noted as context. Draft capital: 2025 R2, pick 60 (Mile High Report/UCF athletics, April 2025) — day-2 capital still buying opportunity in year 2. Pass-pro gate: PFF pass-block grade UNVERIFIED, but the revealed evidence (16% of 3rd-and-5+ snaps as the "lead" back) says the coaches didn't trust him on protection downs — cap the obvious-passing-down projection until camp reports say otherwise. Year-2 leap screen (§11): day-2 capital ✓, late-season snap ≥60% ✓, competition departing ✗ (Dobbins re-signed, Coleman added) — 2 of 3.

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

Offense: 14-3 team, No. 1 AFC seed; top-3 plays/game (66.0 in 2025), +1.6% PROE, 60% neutral pass rate, 20% RB target share. Win total 9.5 → positive script lean. First-time play-caller Davis Webb (Payton oversight, run-game coordinator Strief retained) — system continuity presumed but low-stability; a first-year install drag is priced into team volume.
OL: PFF's No. 1 line of 2025 (6 OL-charged sacks), 4th in RBWR, 5/5 starters returning. Elite environment; also the reason Harvey's bad blocking-adjusted efficiency can't be excused away.
QB: Bo Nix — ankle surgery (Jan + Apr 2026 cleanup), "full go" expected for camp. A Nix setback would drop the whole ecosystem a tier.
Backfield: Dobbins re-signed 2yr/$20M, $8M gtd (March 2026) — real 1A intent money, not lockout money; age 27 with career games 15/0/8/1/13/10 (2020–25; PFR career record, 2024–25 verified against local data). R4 (#108) Jonah Coleman drafted — beat blueprint has him competing with the retained Badie for the third-down/pass-pro role, i.e., attacking Badie's snaps more than Harvey's (Fantasy Life/Heavy, June 2026). Beat consensus: "Dobbins and Harvey in similar roles to last year."
Harvey's offseason: tore his labrum in the AFCCG (Jan 2026), had shoulder surgery, did little team work in OTAs (Denver Post via ProFootballRumors, 2026-06-15), then practiced fully on Day 1 of mandatory minicamp — Payton: "He's out here. Shoot, he's full go now" (SI Broncos, June 2026). Expected full for late-July camp (Denver Sports/DraftSharks, June 2026). SI projects him as "the No. 2 running back behind Dobbins" and the Joker weapon in the passing game (SI Broncos minicamp report, June 2026).

Tripwires (re-run this eval if any fire)

Shoulder setback — Harvey opens camp on PUP or is still limited in August practices → games risk to high, verdict likely drops to HOLD/FADE.
Dobbins injured, traded, or released before/during camp → contingent scenario becomes the base case; likely upgrade toward MUST-HAVE at anything near current price.
Coleman camp reports expand beyond the third-down/Badie role (first-team early-down or goal-line reps) → committee math re-run, ceiling cut.
Beat/coach reports name Dobbins the goal-line back or show Harvey's Joker package shrinking under Webb → the TD and target pillars both weaken; re-run.
ADP rises past ~65 overall (early 6th) → the contingency is no longer free; verdict likely reverts to HOLD.

Sources

data/stats/2025/ + data/stats/2024/: weekly.csv, rushing.csv, receiving.csv, snap_counts.csv, ngs_rushing.csv, participation.csv, pbp_summary.csv (nflverse via nflreadpy, pulled 2026-07-07). Splits, inside-10/inside-5 shares, situational snap shares, success/EPA/breakaway computed this run from raw 2025 pbp (nflverse, downloaded 2026-07-07) joined to participation.csv.
data/adp/adp_ffc_2026-07-07.csv — Harvey 88.4, Dobbins 95.9 (FFC PPR mocks, 2026-07-07).
data/raw/sleeper/players_2026-07-07.json — age/birthdate, college, years_exp.
data/team-profiles/DEN.md (built 2026-07-07) — coaching, OL, win total, backfield contracts/blueprint, team volume inputs.
Web (searched 2026-07-07): ProFootballRumors / Denver Sports / NBC Sports PFT — torn labrum in AFCCG + surgery, limited OTAs (June 2026); SI Broncos — minicamp Day 1 full participation, Payton "full go," No. 2 RB / Joker projection (June 2026); DraftSharks — camp-ready expectation (June 2026); Mile High Report — PFF rookie grades: 68.3 overall / 69.2 rushing, 25 MTF, 2.7 YAC/att (2025 season); Mile High Report / UCF Athletics — 2025 draft R2 #60, UCF production (April 2025).
UNVERIFIED: routes/g and true TPRR (no route-level charting — proxies used and labeled); provider xFP; PFF pass-block grade; Harvey exact career college reception count (NFL receiving sample supersedes); PFR-definition pressure stats.