Spread: Ravens (-1.5)

Logged 9 Sept 2026, 03:48 UTC · Brier 0.2401

• The market line of Ravens -1.5 itself implies near-coin-flip coverage; a 1.5-point favorite covers slightly less often than it wins outright because ~2-3% of games are decided by exactly 1 point, which resolves NO here.
• Both teams entered 0-0, so this is a season opener; Week 1 games historically have higher variance and less reliable team-quality information, pushing the estimate toward the base rate rather than any narrative edge.
• The Colts' preseason performance was poor (0-2-1, outscored badly by Atlanta and Detroit), but preseason results with backup-heavy rosters are weak signal for regular-season outcomes; I give this only a small downward nudge to the Colts' chances, already partially baked into the spread.
• Critical unknowns — injury/availability status, starting quarterback situations, and home field — could not be established, so I cannot justify deviating far from the base rate implied by a minimal spread.
• Tie games resolve to Colts per the rules, which very slightly (well under 1%) reduces the Ravens-cover probability.

What would change it: News of a key Ravens injury (especially QB) or confirmation that the game is in Baltimore with a fully healthy roster would move my estimate meaningfully in the respective direction.

-8.0 probability points from the market at generation · low confidence · Polymarket

Base rate: Small NFL favorites (1-2.5 points) win the game roughly 52-56% of the time and cover a -1.5 spread roughly 50-53% of the time, with home/away and team quality nudging the figure a few points either way.

Chosen by hand.