Bills at Raiders line drops from +6.5 to +4.5; our gap narrows to 5.08
The spread moved from +6.5 to +4.5 on the Raiders while our number barely changed, so the distance between us and the market narrowed to 5.08 points.
Our number
+0.6
Market
-4.5
Gap
+5.1
We publish where our number differs from the market. We do not publish plays. Betting these disagreements is 117-98 (54.4%) forward-graded against 52.4% breakeven, and the matchup model itself covers at 50.9% over 1184 games. A gap is a disagreement, not a recommendation. 21+ where legal.
What changed
The line moved from +6.5 to +4.5 on the Raiders. That is a line_change of -2.0, and it moved toward where we already were. Our own number went from 0.44 to 0.58 over the same window, a change of 0.14. Because the market moved and we mostly stayed put, our disagreement went from 6.94 points to 5.08, an edge_change of -1.86.
The news that came in since our last entry doesn't explain the market's move. Both new items are the same story: CBS NFL and ESPN NFL report that the Rams released kicker Harrison Mevis after his misses in their overtime loss to Buffalo. Our feed tagged these to the Bills and to the field-position phase. Neither club in this game made a roster move, so we treat these items as background, not as an input to Bills at Raiders.
The number
The posted spread is Raiders +4.5 with a total of 53.5. That implies a market margin of -4.5 from the Raiders' side. We project the Raiders by 0.58. The gap is 5.08 points, and our number sits on the LV side.
Part of the difference is structural. Our power ratings spread the league across 7.6 points, while the market's spread across 11.5. We rate Buffalo #6 at 2.29 and Las Vegas #12 at 0.77. That gap between the teams is smaller than the market's, before the matchup is even considered.
Where the game separates
The largest phase gap is ball security: 2.258 standard deviations in favor of LV. The Raiders' offense ranks #9 against a Bills defense at #20 in takeaways vs giveaways. The Bills' offense ranks #29 against a Raiders defense at #7.
Next is pressure: 1.668 standard deviations, also LV. The Raiders' protection ranks #5 against the Bills' pass rush at #22. The other way, the Bills' offense ranks #26 against the Raiders' rush at #10.
The third is hidden yardage (field position): 0.923 standard deviations, LV. The Raiders' offense ranks #1 against the Bills' #13. The Bills' offense ranks #6 against the Raiders' defense at #7, so the gap is smaller on that side.
These phases describe the matchup. They are not a forecast of the result. Over a sample of 1184 games, our matchup model's verdict is: "Does not beat the spread. The quartile it is most confident about is its worst." It is at 50.9% against the spread overall and 47.3% in its top quartile. A game can separate on paper without that telling us anything about the cover.
On the wire
- CBS NFL has a status piece on DJ Moore, whom it lists as a Bills wide receiver, ahead of a Monday night game against the Rams. Our feed links it to explosive plays. The item is about practice status, and nothing in it is confirmed. - CBS NFL and ESPN NFL report the Mevis release described above. The phase tag is field position, but the move belongs to the Rams. - Rotowire reports Blake Corum was "efficient with eight touches." It touches no phase in our model.
We are watching for real practice or injury reports involving Bills or Raiders players. None of the current items is one.
What would change our mind
We will write again if any of these happen:
- The spread moves 1.0 point from +4.5, or crosses 3 or 7. - The total moves 1.5 from 53.5. - Injury news comes in that affects a starter in the ball-security, pressure or field-position phases above. - Our own rating for either team moves 1.0 after this week's games are folded in.
Until then, the record stands as written: we have the Raiders by 0.58, the market has them getting 4.5, and the two numbers are 5.08 apart. 21+ where legal.
Bills-Raiders moves to +6.5 and our gap to the market widens to 6.94
What changed
Since our last entry, the spread has moved a full point against the Raiders, from +5.5 to +6.5. Our projection did not move with it. Our gap to the market went from 5.94 points to 6.94, so the whole 1.0-point move landed in the disagreement.
Two new Rotowire items came in during that window: "Blake Corum: Efficient with eight touches" and "Davante Adams: Two TDs in OT loss." Our feed tags both to BUF. Neither maps to a phase in our model, and neither changed our rating.
The number
Las Vegas is +6.5 at home with the total posted at 53.5. Kickoff is Sun 4:25PM ET. The market's projected margin from the Raiders' side is -6.5. Ours is 0.44, close to a coin flip. That puts us 6.94 points off the market, on the Raiders side.
In our power ratings the Bills rank #6 (2.43) and the Raiders rank #12 (0.77). One piece of context belongs next to that: our ratings span 7.6 points across the league, while the market's span 11.5. Our model separates teams less than the market does. Part of a gap this size between a #6 and a #12 team comes from that compression, not from anything specific to this game.
Where the game separates
All three of the top phases lean LV.
Ball security (takeaways vs giveaways), 1.597 SD. The Bills offense ranks #29 here and faces a Raiders defense at #7. On the other side, the Raiders offense is #9 against a Bills defense at #9.
Pressure (pass rush vs protection), 1.256 SD. Buffalo's offense ranks #27 in protection against a Raiders pass rush at #10. The Raiders offense is #5 against a Bills defense at #11.
Hidden yardage (field position), 1.092 SD. The Raiders offense ranks #1 against a Bills defense at #13. The Bills offense is #9 against a Raiders defense at #7.
These phases show where the matchup is lopsided on paper. They do not predict who covers. The matchup model behind them is 50.9% ATS over a sample of 1184, and its most-confident quartile is at 47.3%. Its own verdict: "Does not beat the spread. The quartile it is most confident about is its worst." So the Raiders' edge in the trenches and in ball security describes how this game could play out. It does not support the side. The idea we are testing next is whether phase separation adds any information once the power rating is already accounted for.
On the wire
ProFootballTalk reports that Bills WR DJ Moore and CB Dee Alford were inactive for Monday Night Football against the Rams. CBS NFL ran a separate look at Moore's status for that game. Both reports are about Monday's game, not this one. We have no report yet on Moore's status for Sunday. In our model his item touches explosive-play rate, which is not one of the three phases above.
The Rotowire notes on Corum and Adams are tagged to BUF but carry no phase. Two CBS NFL items in the feed are sportsbook promotions with no information about the game, and we ignore them.
What would change our mind
We write again if any of these happen:
- The spread moves 1.0 point from +6.5, or crosses 3 or 7. - The total moves 1.5 from 53.5. - Injury news that gates a player, such as a confirmed Sunday status for Moore. - Our own rating for either team moves 1.0 after this week's games.
We publish a number and where it differs from the market. This is not betting advice. 21+ where legal.
Bills rule out DJ Moore, Dee Alford for Monday night; total moves to 53.5
What changed
ProFootballTalk reports that Bills WR DJ Moore and CB Dee Alford are inactive for Monday Night Football against the Rams. ProFootballTalk also reports that Buffalo elevated WR Trent Sherfield and DL Phidarian Mathis for that game, and CBS NFL ran a status check on Moore before kickoff. Our board maps all three items to the explosive-rate phase.
These reports are about Monday's game, not Sunday's trip to Las Vegas. Nothing on the wire yet says whether either player is available for this one. This update also shows no change to our projected margin. The total did move. It was 52.5 when we last wrote, it is 53.5 now, and that is a move of 1.0.
The number
We project the Raiders' margin at 0.44. The market's projected margin is -5.5, from a posted line of 5.5 with Buffalo favored. That puts us 5.94 points from the market, on the Las Vegas side.
Some of that gap comes from how the two scales are built. Our ratings span 7.6 points from top to bottom, and the market's span 11.5. Buffalo is #6 in our power ratings at 2.43. Las Vegas is #12 at 0.77. On our narrower scale, the difference between those two teams is smaller than the market's spread implies.
Where the game separates
The three widest phase gaps all lean Las Vegas.
Ball security (takeaways vs giveaways): 1.597 SD. The Raiders' offense ranks #9 against a Bills defense that also ranks #9. The bigger mismatch is on the other side: Buffalo's offense ranks #29 against a Las Vegas defense at #7.
Pressure (pass rush vs protection): 1.256 SD. The Raiders' offense ranks #5 against Buffalo's #11 defense. Buffalo's offense ranks #27 against a Las Vegas defense at #10.
Hidden yardage (field position): 1.092 SD. Las Vegas has the #1 offense in this phase, facing a Bills defense at #13. Buffalo's offense ranks #9 against the Raiders' #7 defense.
This describes the matchup. It does not predict the result. Our matchup model is 50.9% against the spread over 1,184 games, and its most confident quartile is 47.3%. The model's own verdict: it does not beat the spread, and the quartile it is most confident about is its worst. So the separation shows where the game should be decided. It does not tell you who covers.
On the wire
- ProFootballTalk: Moore and Alford are inactive for Monday night against the Rams. Touches explosive rate. - CBS NFL: a status check on whether Moore would play Monday. Touches explosive rate. - ProFootballTalk: the Bills elevated Sherfield and Mathis for Monday night. Touches explosive rate. - CBS NFL: two sportsbook promotion pieces tied to the Monday game. They carry no football information and touch no phase.
Explosive rate is not one of the three phases where this game separates most. Even if Moore's absence carries into Sunday, it would affect a part of the matchup outside where our gap sits. Alford is a corner, so his status matters more for how Buffalo defends the Raiders' passing game. Neither report tells us anything about Sunday yet.
What would change our mind
We will write again if any of these happen:
- The line moves 1.0 point from 5.5, or crosses 3 or 7. - The total moves 1.5 from where it sits now. Today's move of 1.0 is short of that. - Gating injury news arrives, especially a confirmed status for Moore or Alford for Sunday. - Our own rating for either team moves 1.0 after this week's games, including Buffalo's Monday result.
This is a published number and where it differs from the market, not a recommendation. 21+ where legal.
BUF at LV: our number sits 5.94 points away from the market's -5.5
The number
The Bills visit the Raiders on Sun 4:25PM ET. This opens our thread for 2026-W42.
The posted line has LV at 5.5 and the total at 52.5. The market's projected margin for LV is -5.5. Ours is 0.44. That leaves 5.94 points between us, and our number is on the LV side.
Part of the gap comes from how the two systems rate teams. Our power ratings have BUF at #6 with a rating of 2.43 and LV at #12 with a rating of 0.77. That's a real difference, but it's smaller than the market's. Across the league our ratings span 7.6 points and the market's span 11.5. We spread teams less than the market does, so a favorite of this size will often look smaller to us. Keep that in mind when reading the 5.94.
Where the game separates
Our matchup model shows the same three phases at the top, and LV has the edge in all of them. This is a description of the matchup, not a forecast. The matchup model is 50.9% against the spread over a sample of 1184. In its most confident quartile it is 47.3%. Its verdict: "Does not beat the spread. The quartile it is most confident about is its worst." So the separation below tells you what kind of game this is. It does not tell you who covers.
Ball security (takeaways vs giveaways): 1.597 SD, LV. This is the widest gap on the card. LV's offense ranks #9 against BUF's defense at #9, which is even. The difference is on the other side: BUF's offense ranks #29 in this phase against an LV defense at #7.
Pressure (pass rush vs protection): 1.256 SD, LV. It's the same shape. LV's offense ranks #5 against BUF's defense at #11. BUF's offense ranks #27 against LV's defense at #10. In both of these phases, the BUF offense is where the matchup tilts.
Hidden yardage (field position): 1.092 SD, LV. LV's offense ranks #1 here, against BUF's defense at #13. BUF's offense at #9 against LV's defense at #7 is closer to even.
These are LV's three largest edges, and they line up with where our number lands. None of them is the kind of phase the model has shown it can price against a spread.
On the wire
Rotowire NFL reports that TreVeyon Henderson is headed for testing Monday. Our wire tags the item to LV and doesn't map it to any phase.
CBS NFL has a look at how Fernando Mendoza played in his NFL debut. It's the only wire item tied to a phase: dropback efficiency. That phase is not one of the three above, so it isn't driving the separation, but it's the item most likely to matter to the LV offense.
ProFootballTalk's recap of the Patriots' win over the Raiders is tagged to LV with no phase attached. On the BUF side, CBS NFL has a piece on Matthew Stafford's season ahead of the Bills' game against the Rams. We note it as context and attach no phase. A promotional item tagged to BUF has no football content and plays no part here.
What would change our mind
We'll write again on any of these:
- the line moves 1.0 point, or crosses 3 or 7 - the total moves 1.5 from 52.5 - injury or depth-chart news that gates a starter, especially at quarterback for either side - our own rating for either team moves 1.0 after the week's games
Until one of those happens, this is where our number sits. It's 0.44 against -5.5. We aren't telling anyone how to act on it. 21+ where legal.
This thread updates when the line moves a point or crosses 3 or 7, when the total moves 1.5, when gating injury news lands, or when our own rating moves after the week's games. All matchups.