NFLNO PLAY research

Sam Roush

Position and team not on the research feed

Next gameChicago Bears @ Seattle SeahawksMon, Nov 2, 8:15 PM ETGame intelligence →

Historical sample Measured: 5 games
Last 10 · receptions Measured: —
MCP consistency Model: —
Markets evaluated: 4

Evidence · receptionsNO PLAY

Line Market 0.5

Price-210

Sportsbook—

REPICOM modelModel59.6%

Market impliedMarket67.7%

Model − market-8.1 ptsPrice -210

Edge Derived: -8.1 pts
EV per 1u Derived: -0.120u
Sample Measured: 5 g

REPICOM decisionNO PLAY

Model signal exists, but the player does not meet the 20-game publication history floor.

A positive gap is not a pick: qualification also requires confidence, edge, expected value and a sufficient sample — those rules are unchanged.

Source & method · as of Nov 1, 3:32 PM ET

Source: repicom_public_nfl_intelligence_v1 (NO PLAY research rows, read server-side)
Statistic: Probability of the evaluated side, in %
Calculation: Model = model_probability. Market = captured market_probability; if absent, implied from american_odds (−odds ÷ (−odds + 100), or 100 ÷ (odds + 100)). Gap = model − market, in percentage points. EV = p × (decimal − 1) − (1 − p) per 1u.
Sample: history_games on the row
Freshness: Read on every request (no cache); the market price is the one captured with the evaluation
Missing data: A missing probability or price shows “—” and no bar. Nothing is estimated.

Recent form · receptionsMeasured

Recent formNot available yet — requires a verified non-qualified intelligence row for this player and market. REPICOM does not estimate missing data.

Source & method · as of Nov 1, 3:32 PM ET

Source: repicom_nfl_intelligence_v1 — non-qualified, unlocked rows only, joined on player + market (+ game_id)
Statistic: Games that cleared the line, out of the last 10 (count)
Calculation: recent10_hits ÷ recent10_games. Dots show the count, not the game order.
Sample: recent10_games
Freshness: Read on every request (no cache)
Missing data: No confident join → “Recent form unavailable”.

Market map · every evaluated marketTap a market

REPICOM modelMarket impliedGap in pts

Compare playersPublic research only

Compare with

Sam Roush

Rome Odunze

Historical sample

5 g9 g

Last-10 hit rate

—89%

Best model − market gap

+3.3 pts+17.6 pts

Markets evaluated

Open Rome Odunze

Game-by-game · real game logs vs the Hard Rock line

ReceptionsNo Hard Rock lines in these games

2026 season: 0.6
Last 5: 0.6
Last 10: 0.6
Home · Away: 1.0 · 0.0

MIN

PHI

NYJ

ATL

Date	Opp	Receptions	Line
Oct 18	@ ATL	0	—
Oct 11	@ GB	0	—
Oct 4	vs NYJ	1	—
Sep 28	vs PHI	1	—
Sep 20	vs MIN	1	—

Receiving yardsNo Hard Rock lines in these games

2026 season: 6.6
Last 5: 6.6
Last 10: 6.6
Home · Away: 11.0 · 0.0

MIN

PHI

NYJ

ATL

Date	Opp	Receiving yards	Line
Oct 18	@ ATL	0	—
Oct 11	@ GB	0	—
Oct 4	vs NYJ	8	—
Sep 28	vs PHI	16	—
Sep 20	vs MIN	9	—

Stats: nflverse game logs, regular season. Line: Hard Rock's final line before each game, when we captured one. Green bar = over the line, red = under.

Also researched in this game8 players

Rome Odunzereceiving yards8/9 last 10 Caleb Williamspassing yards5/6 last 10 AJ Barnerreceptions7/10 last 10 Cooper Kuppreceiving yards6/10 last 10 Jaxon Smith-Njigbareceiving yards10 g sample Luther Burden IIIreceptions6/9 last 10 D'Andre Swiftrushing yards36 g sample Colston Lovelandreceptions9 g sample

SourcesResearch: repicom_public_nfl_intelligence_v1 (NO PLAY rows) · Form: repicom_nfl_intelligence_v1 (non-qualified rows only) · Read live at Nov 1, 3:32 PM ET.Member-qualified and locked player markets are never shown on public research pages. A research signal is not a REPICOM play.