Picking — each entry's cell holds both teams; tap one to take that side. The system's recommended side is marked with a dot, but either side is selectable: a coin-flip game is still a coin flip. Every other column on the row describes the SYSTEM PICK side.

POOL LINE is the CBS pool's posted spread — the number the pool scores against, from the pick side's perspective. MARKET is the current market line for the same side — the median of Pickwatch, the sportsbook line CBS carries, and nflverse. Hover it to see how many sources agreed. The gap between the two is the stale-line edge.

1 · Stale Line — how many points better your pool number is than the current market number. This is the system's main edge: the pool scores you against a line the market has since moved away from. gains the 3 / 7 appears when the move carried your side across 3 or 7 — the two most common final margins in the NFL — which is worth more than the raw points suggest.

2 · Public Bias — which side the crowd is on and how heavily, named from the side they are actually on. The pool takes favourites about 65% of the time, so heavy chalk and fading the favourite mark games where it is behaving unusually.

3 · Ownership — the share of each crowd holding the system pick. cbs and espn are observed; j4k is modelled for Just4Kicks specifically.

COVER PROB — the chance this side beats the pool's spread: how far the market has moved off the pool number and which final margins the move crossed, plus a flat 2.5-point calibration because the raw edge runs that much low over 2,112 past movers. Nothing about the teams, the division or the time slot is in it — the old divisional and favourite adjustments were retired in week 2 of 2026 when a fuller test showed the favourite one had the wrong sign (favourites cover 48% against the pool line) and the divisional one did not hold up. Values marked * are for games whose line has not moved, where there is no edge to measure: 55.5 with four or more signals, otherwise 51 for the underdog and 49 for the favourite, which is the population lean and is there to order the fills, not to be believed to the decimal. Colour: green ≥58, blue ≥54, amber ≥50.

QB OUT — that team has a quarterback listed Out or Doubtful on the nflverse injury report (updates Wed–Fri). In a 2018–24 retro-test, movers where the line moved against the team missing its QB covered 67% (n=94) vs 59% for other movers; movers where the line moved toward the team missing its QB covered 38% (n=42) — shown in red. Small samples; context, not a score.

LINE NOTE — one sentence on why the market has moved off the pool line (injury, weather, reported sharp or public action, news). Refreshed every evening, and again through Sunday as teams announce who is playing. Reported, not verified; never a recommendation. Every player named is checked against the real roster, and a note that puts a player on the wrong team is thrown away rather than shown. “Line moved X since” means the market has moved since the note was last checked, so the note may not explain the number you are looking at now. Click a note to edit it. The tag is kept so games can be grouped by cause after the season.

Ordering — Tier 1 is every moved line that crossed a 3 or a 7 (62% over 632 past games); Tier 2 is every other moved line (56% over 1,480). Inside a tier, games rank by cover probability, and the side fewer people are on breaks ties. Below the movers there is no line move to read, so the fills rank on pick value, which puts four-signal games first, then underdogs, then favourites, each group ordered by lowest ownership. The ownership tiebreak is free: measured across four separate readings of the crowd (our own pool, CBS national, ESPN, DraftKings tickets), ownership on games without a line move has no relationship with covering. The underdog lean is small (51% vs 49%) but it is the direction the crowd's favourite bias points away from.

STANDING — expected places gained in a 211-entry pool if you take this side: you pass everyone who doesn't have it when it covers, and fall behind everyone who does when it doesn't. This is the "high cover probability, low ownership" trade as one number, and it needs no fudge factor — it is just pool size × (cover probability − ownership). Read it as context, not a ranking. It is a per-pick figure and overstates season value: ownership doesn't change your expected wins, only how closely your score tracks the field's. Ordering picks purely by this scored 56.6 wins a season against 62.9 for ordering by cover probability.

AGREE — of the 19 backtested strategies, how many select this side, split by strategy family. Hover for names.