The model gives Buffalo a small edge on the road, projecting a 2.9-point Bills win against a market that has Houston as 1-point favourites. The gap versus SportsBet's -0.5 line for Buffalo (best AU price at 1.9) is only 1.4 points, well short of the 10-point bar the model needs before week 6. The total is the primary market here: the model projects 42.1 against a line of 44.5, a 1.9-point edge at SportsBet's 1.9 price on the under, again short of threshold. Both markets are labelled PASS, so this is analysis, not a tip. The early-season record: 32-46 against the spread and 41-37 on totals in weeks 1 to 5 last season, with totals at that rate essentially break-even.

There's no 2026 play-by-play yet, so both ratings blend last season's form with the market's post-offseason view of each team (the market anchor), scaled by roster continuity. The one live signal is Buffalo's short-left passing game, rated +0.23 EPA/play better than Houston's defence handles it, part of why the model leans Bills, but with Houston's defence otherwise sound and no quarterback hand adjustment applied, nothing here clears the model's own bar.