Malik Willis

Malik Willis — QB, MIA (2026)

Verdict

TARGET (medium confidence) at a free price — mock-undrafted in 1QB. Willis is a $45M-guaranteed Week 1 starter (ESPN, 2026-03-09; Dolphins on SI, June 2026: "zero" doubt he opens as QB1) with genuine dual-threat traits, elite college rushing pedigree, and a career 4.9 YPC on 76 NFL carries — attached to the league's worst passing environment and a live mid-season leash. Why the market is wrong: the 1QB market prices him as a non-entity (no ADP), but a locked-in Week 1 starter with a plausible 5–6 carries/game and 4–5 rush TDs has a QB1-spike-week ceiling at literally zero acquisition cost; that asymmetry is +EV for anyone punting QB, streaming, or needing a bye-week/upside QB2. The bet is on the rushing floor and the contract leash, not the passing. If camp shows no designed-run package, the thesis dies (tripwire 2) — this is a dart, not a starter plan.

Projection & comps

Bottom-up, two components (scoring-framework §2), from the MIA team profile environment (~60.5 plays/gm, ~28 pass att/gm, win total 4.5, bottom-10 OL — data/team-profiles/MIA.md, 2026-07-07):

Passing (median, 15 starts): 28 att/gm × 6.6 YPA (environment-derived: thinnest WR room in the league, rookie LG, YAC-scheme mitigation; his 10.2/12.1 GB YPA on 89 career attempts is scheme-and-sample artifact, discarded) ≈ 2,770 yds. Pass TDs anchored to xTD logic, not his GB rate: ~3.6% on 420 att ≈ 15 (a 4.5-win team is red-zone poor and the QB-run threat siphons goal-line TDs). INTs from TWP reasoning, not his clean ledger: career 0 INT on 89 GB att is luck + a protected script (his TEN sample was 3 INT on 66 att; MIA beat writers flag processing speed) → ~2.5% ≈ 10–11.

Rushing (projected separately — it is the floor): 6.0 carries/gm (≈3 designed + ≈3 scrambles; GB 2025 = 5.5/gm across 4 games, nflverse; Sun-Sentinel/Perkins 2026-07-04 pegs the sensible target at 5/gm) × 5.2 YPC ≈ 31 yds/gm → ~470 yds. Rush TDs anchored to a projected xTD of ~4 (partial sneak/keeper role; Ollie Gordon contests goal-line — team profile). Rushing alone ≈ 4.5–5.0 PPG of weekly floor.

Games risk: high — a ~6 carry/gm runner behind a bottom-10 OL (qb.md §3 games-risk nudge) *plus* benching exposure (Ewers). No provider xFP available — usage-based estimate only, marked as such.

Comps: Marcus Mariota 2022 ATL (bridge dual-threat on a run-committed team, ~15.5 FPG, benched in December — the median-with-leash script) · Justin Fields 2023 CHI / 2025 NYJ (rushing floor carrying a broken passing offense) · Tyrod Taylor 2015 BUF (efficient bridge dual-threat, top-10 finish — the ceiling) · Anthony Richardson 2024 IND (accuracy crisis + benching drama on a run-heavy offense — the floor).

No external projections on hand (data/projections/ does not exist) — no sanity-check disagreement to log.

Usage profile (qb.md §2 opportunity core — projected 2026, with 2024–25 GB evidence)

Efficiency/talent (qb.md §5): career sample is 11 GB games / ~130 dropbacks — below every minimum. GB 2024–25: 70/89 (78.7%), 972 yds, 6 TD, 0 INT; 2024 led the NFL in YPA (10.19) among 50+ att passers (packers.com/NFL.com via search, 2026-07-08). Weekly CPOE in GB work was strongly positive (e.g., +21.3 in the Wk17 2025 BAL start, 18/21 for 288 — nflverse ngs_passing/weekly.csv) but on protected, run-first scripts. Contra-signals: TEN 2022–23 was 53.0% comp, 0 TD/3 INT (49.4 rating); 2024 sack rate ~13% (8 sacks on ~62 dropbacks — nflverse passing.csv) says pressure-to-sack is a live concern behind a bottom-10 OL; TWP UNVERIFIED (no charting provider). The 0-INT ledger over a tiny, schemed sample is exactly the "luck about to land" pattern (qb.md §12) — priced in here only because the price is zero.

Pedigree (prospect-pedigree.md): 2022 R3 #86 (day-2 — bridge-timeline capital, now year 5: capital prior fully decayed, believe the NFL record). College: Liberty 2020–21 — 944/14 and 878/13 rushing seasons (Wikipedia, fetched 2026-07-08); elite college rushing production, the QB trait that translates best (§2). G5 competition + no combine 40/testing (RAS unavailable — UNVERIFIED), but NFL rushing tape/production confirm the athleticism, which is the only testing QB evals care about (§3). Age 27 (b. 1999-05-25, Sleeper 2026-07-07) — pre-rushing-decline window; no age haircut yet, first haircut year would be 2027 (age 28).

Context (team profile: data/team-profiles/MIA.md, built 2026-07-07)

• Total teardown, stability: low. New GM (Sullivan), new HC (Hafley), new play-caller (Slowik, Shanahan tree — zone/PA/RPO, 21-personnel heavy), QB1 and the entire WR1/WR2 tier gone (Tua released, Hill released, Waddle traded). Year-1 install drag applies.
• Job: 3yr/$67.5M, $45M gtd over two years — the money wins camp ties. Backup Quinn Ewers (2025 R7, 3 late-2025 starts: 66.7%, 569 yds, 3 TD/3 INT — NFL.com/heavy.com) outplayed expectations at minicamp ("Ewers isn't competing to be Dolphins' starter, but maybe he should" — Omar Kelly via heavy.com, 2026-06-05) and knows the system. Per qb.md §8 the benching triad is only ~1-of-3 (low backup capital, regime paid Willis) → Week 1 secure, mid-season leash real.
• Usage intent — the swing variable: Slowik calls Willis's legs "definitely a weapon" that "lets you equalize" (NFL.com/PFT, June 2026) but the staff's stated ideal is "an agile pocket-passing quarterback who has the ability to run" (Sun-Sentinel/Perkins, 2026-07-04), and OTA observers saw few runs and a deliberate pocket-development emphasis. No Slowik designed-QB-run history exists to project from.
• Environment: worst-in-class — vacated ~46% of targets replaced by day-3/cheap vets; Achane is the offense; OL bottom-third in both win rates with a rookie LG; Vegas win total 4.5 with the second-hardest schedule → heavy trailing scripts that add garbage-time dropbacks but subtract red-zone trips (qb.md §7 trap: bad team without a competent passing game is not the garbage-time-hero screen).

Tripwires (re-run this eval if any fire)

• Ewers takes first-team reps into August or is named Week 1 starter → verdict void, AVOID at any price.
• Camp/preseason shows no designed-run or red-zone keeper package (or Slowik publicly rules it out) → downgrade to HOLD/AVOID; the thesis is the legs.
• Willis benched in-season → drop; Ewers becomes the (non-rostered) MIA QB.
• 1QB ADP climbs inside ~pick 180 / QB20 on camp hype → price no longer free; re-verdict (likely HOLD).
• MIA adds a veteran QB or a real WR1 (e.g., trade/signing with starter capital) → re-run team profile and this eval, either direction.