AoA
Join
← Intel
How the number is built · 5 min read

How an NFL point spread is actually built

Power ratings + home field + the situational components — with the real numbers.

Spread = (Home Power Rating + Home-Field Advantage) − Away Power Rating. The rating gap is the base; the situational components are adjustments on top.

Home field went DOWN, not up

Historically ~3.0 points, now ~2.0 and falling. Home teams averaged +2.33 margin from 2010–2018, but only +0.54 since 2019. Current consensus for 2025–26 is ~2.0 (some models 1.5). Most fans still think it's 3.

QB injury is the biggest single line-mover

Only a starting-QB injury (or a few elite skill players) moves the spread meaningfully. Oddsmakers value Josh Allen at ≈ 6.98 points and Mahomes at ≈ 6.94 — an elite QB is worth nearly a full touchdown. A starting-QB injury can swing a line 6–7 points. There is no bigger single factor in football.

Weather — real and quantified

  • Wind (biggest): negligible under 15 mph; 20+ mph ≈ 2.7 fewer points, completions drop ~10%.
  • Cold: below 25°F, ~8% fewer points; denser ball → shorter kicks.
  • Rain: passing down ~12%; heavy rain up to 6 fewer points.
  • Snow: heavy snow ~25% fewer points; FG conversion falls to ~76%.

The honest rail — is it already priced?

Mostly yes. Books open lines with weather and injury models baked in and adjust as forecasts and statuses firm up. So the edge is NOT 'weather/injury matter' (the book knows). The edge IS in timing (catching the move before the line firms) and magnitude (when the market over- or under-reacts to the news vs. its true point value).

We don't predict the game — we show you how the number is constructed and where the crowd misjudges a piece of it.

Source · Adapted from our Anatomy of a Point Spread research · theScore, nfelo, Yahoo, Covers, Pinnacle

Informational decision-math education, not betting advice. 18+. SIM-safe · read-only odds. We show our calibration and our losers.