Buffalo Bills at Houston Texans

Where the fixture comes from. Who hosts is not in the Scheme Score export. It is read from the prop capture, which marks an away team, and cross-checked against the FantasyPoints matchup cards, which title themselves the same way. The two agree on 16 of 16 games. Neither source carries a venue, so none is shown.

Where the price comes from. From one dated prop capture, the same one the Scheme Score export now carries on its own priced rows, so the feed and this page state one answer rather than two. A line here is the consensus across independent sources in that capture, joined to the board by player name and team, and the two prices are the median quote at that line. A row with no line was quoted by nobody in the capture under the name and team the board uses; that is unpriced, not a line of zero. The Sources column says how many sources stood behind a number; it never says which ones, and no operator is named anywhere on this site. The capture is dated 2026-09-09 and is recorded to the day, not to a clock time, so no reader may date it more finely. One capture, so there is no line movement and no closing line.

Availability is not stated here. The 2026 week 1 practice report on this desk carries eleven rows across two of the thirty-two teams, and not one of them states a report status. Official inactives are published in no source here. So these pages say nothing about who plays, and a player appearing on them is not evidence that he will.

Scheme Score is a matchup feature built for player-prop gambling research. It is one input to a betting decision, not a standalone player projection, hit probability, expected value estimate, or stake size. Most rows on this board now carry a grade. The man and zone coverage signal is live: both the man/zone fit and the usage fit legs contribute to a grade, reflecting whether a receiver's style and usage lean toward the coverage this defense plays most. The slot coverage funnel is a direct measurement from PFF's slot coverage report, not an estimate. Some inputs are still missing. Direct TE and RB receiving funnels are not available: tight ends run through the linebacker coverage proxy, and running backs get no receiving funnel signal at all. Defender coverage grades still do not contribute to any grade, because that leg needs each defense's 2026 returning-share data, which does not exist here. The defensive coordinator table is an unfilled stub, so no coordinator discount is applied to any row. A small number of rows stay ungraded because the player name on that row is ambiguous and the grader skips rather than guesses. This board is two tiers. 134 of 290 rows are priced: they carry a posted receiving-yards line from one dated prop capture, joined to the board by player name and team, plus the median over and under price quoted at that line and a count of how many independent sources stood behind it. The other 156 rows carry no line, which means no source in that capture quoted them under the name and team this board uses; that is unpriced, not a line of zero, and it is not a claim that the player will not play. Which operator quoted a number is withheld everywhere in this feed; how many stood behind it is published instead. The capture is recorded to the day and not to a clock time, so no reader may date it more finely, and it is a single capture, so there is no line movement and no closing line here. The line does not enter Scheme Score: the grade is computed from prior-season scheme and usage evidence alone, so a grade is never a projection of the line beside it.