The model gives Bills a 7.3-point edge over Lions, but against the market's best line of BUF -4.5 at Dabble AU (1.9) that only translates to a 2.8-point gap, noted but short of the Lean floor. This early in the season the rating still leans heavily on last year's numbers blended with the market's post-offseason view of both teams and scaled by roster continuity, with the 2026 sample so far just 1,830 plays from week one and still mostly outweighed by that prior-season and market anchor. Buffalo's deep-right passing rates at +0.56 EPA/play against Detroit's defence, the weakest zone in this matchup, but that's a model rating rather than a reason to treat this as more than an edge noted. The model's early-season record sits at 1-0 against the spread.

Secondary market interest is the total: model projects 52.6 against a market line of 55.0 (PlayUp, 1.85) for an edge of 1.9, which is a PASS. Detroit's inside runs graded at +0.19 EPA/play against Buffalo, one of the few positives for the visitors, while on last season's charting Josh Allen dropped to -0.518 EPA per dropback under pressure, both context rather than reasons to move off the model's numbers here.