American Football Intelligence · NFL 2026 · Regular season · Week 5
Bills @ Rams
Monday, October 12 · 5:15 p.m. PDT / 8:15 p.m. EDTSoFi Stadium · Inglewood · October 13, 00:15 UTC · Non-neutral site
Four accepted 2026 games per team. No selected-game outcome or postgame rating is available before kickoff.
Authoritative launch markets
The prices being evaluated
Draft Kings · All selections: full_game · espn_web / ESPN source [https://site.api.espn.com/apis/site/v2/sports/football/nfl/scoreboard?dates=20261012&limit=1000]Retrieved . Bookmaker quote-update time not supplied; overtime inclusion unknown.
| Market | Selection | Signed line | American price | Price break-even |
|---|---|---|---|---|
| Moneyline | Los Angeles Rams | — | -180 | 64.29% |
| Moneyline | Buffalo Bills | — | +150 | 40.00% |
| Spread | Los Angeles Rams | -3.5 | -105 | 51.22% |
| Spread | Buffalo Bills | +3.5 | -115 | 53.49% |
| Total | Over | 54.5 | -115 | 53.49% |
| Total | Under | 54.5 | -105 | 51.22% |
Break-even is the reciprocal of the supplied decimal price, not an estimated outcome probability; unknown settlement rules remain unresolved. The supplied launch snapshot is retained. Historical benchmark market lines are not these prices.
Betting decision
PASS pregame · the evidence does not establish a dependable price edge
Accepted forecasts generally lean toward a Rams win, but the best margin benchmark points to a close game. The latest descriptive SRS favors Buffalo, the total models disagree by 23.17 points, and important availability decisions remain pending.
Rams win direction
-180 needs 64.29%. AMF Power estimates 65.63%; Margin Elo estimates only 54.87%.
Bills +3.5 (-115) lean
A three-point loss covers. The 53.49% cover threshold cannot be verified from mean margins.
Under 54.5 (-105) lean
Predictive SRS supports a lower mean total, but two other accepted models support much higher scoring.
Betting conviction: low. These are qualitative directions, not validated cover/total probabilities, expected-value estimates or recommended stakes.
Canonical evidence · 2026 only
Each offense meets the opposing defense
Season aggregates: four complete games each; no partial or box-only games. Metrics as of October 4, 17:00 UTC; calculated October 12, 07:01:19 UTC. Ranks shown only where supplied. Lower defensive EPA, success and explosive rates allowed are better.
Buffalo offense → Los Angeles defense
Samples: Buffalo 235 offensive plays / 43 drives; Rams 239 defensive plays / 43 drives.
Allen's passing EPA is +0.265 per recorded dropback (128 opportunities, four games). Cook has 421 rushing yards and 15 explosive carries on 70 attempts; Allen adds 33 carries and seven rushing touchdowns. Los Angeles' strong overall defensive profile is a substantial obstacle. Accepted pass/rush-specific defensive splits are unavailable in the season summary, so neither route is assigned a numerical matchup advantage.
Los Angeles offense → Buffalo defense
Samples: Rams 268 offensive plays / 47 drives; Buffalo 271 defensive plays / 43 drives.
Stafford has 1,189 passing yards, six touchdowns and six interceptions across four games; passing EPA is +0.080 per recorded dropback (168 opportunities). Adams has 390 receiving yards on 38 targets, and Williams has 294 rushing yards on 54 carries. Buffalo's permissive aggregate defense offers an opening, but Rousseau's six sacks, 16 QB hits and three forced fumbles across four games are a credible disruption threat.
Player metrics: football-metrics-v1, selected-season source-native aggregates. Above offensive and Rousseau samples have no declared missing baseline fields; defensive pressure counts are unavailable. Passing opportunities are the database's recorded QB measure and do not establish called-play tendencies.
Traditional SRS · descriptive anchor
The rating cutoff changes the ordering
Latest season values · October 12, 00:20 UTC
Buffalo leads the current descriptive comparison. Its +5.00 average scoring margin compares with Los Angeles' +0.50.
2026: four team games each; complete coverage; 78 league-network input games. traditional-srs-v1. Rating uncertainty 5.11 is descriptive, not a game-margin distribution.
Fixed matchup snapshots · prior periods only
The accepted game snapshot favors Los Angeles by 2.78 rating points before the model's location treatment. Its declared forecast is Rams +4.98, with a 68.25% model win estimate.
Training cutoff October 9, 00:15 UTC; latest training kickoff October 6, 00:15 UTC; 64 league games, 32-team connected component, four games per team. Pregame ranks not supplied. Descriptive uncertainty 4.45.
Style, recent form and confidence
Explosiveness and lineup changes make four-game averages fragile
Buffalo: multiple routes, uneven recent ball security
The offense's 47.7% success rate is close to the Rams' 47.0%, but Buffalo's explosive rate is much higher: 15.3% versus 10.1%. That suggests larger gains contribute materially to its scoring advantage; it does not prove greater game-to-game variance.
Cook's rushing production and Allen's rushing touchdowns give Buffalo a way to attack without relying entirely on a depleted receiver group.
External 2026 context: the Rams' team preview reports no Buffalo giveaways in the first two games and seven in the next two, plus multiple-tight-end usage on more than half its plays. Those recent swings lower confidence in extrapolating 31.75 points per game. October 9 team preview [https://www.therams.com/news/all-eyes-on-sofi-stadium-rams-bills-battle-week-5-monday-night-football-game-preview]
Los Angeles: returning personnel changes the offensive sample
Nacua's accepted receiving sample is only two games: 14 catches, 199 yards and 21 targets. It includes his return after missing two games. A season average with and without him is a mixed lineup sample; it is not a stable estimate of this week's offense. October 9 Nacua availability context [https://www.therams.com/news/feels-good-to-get-my-feet-truly-under-me-wide-receiver-puka-nacua-shines-against-eagles-prepares-to-replicate-2024-performance-against-bills-on-monday-night-football]
The team preview reports multiple giveaways in all four games and only one three-tight-end snap against Philadelphia with Parkinson and Ferguson unavailable. Parkinson's expected return could expand those formations. 2026 personnel and ball-security context [https://www.therams.com/news/all-eyes-on-sofi-stadium-rams-bills-battle-week-5-monday-night-football-game-preview]
External research · official team reports
Availability alters both sides of the matchup
Official final practice reports published October 10; observed October 12 during 08:44–08:49 UTC research. Pregame inactives have not been verified. “No designation” is clearance to be available, not confirmation of snaps or workload.
| Team / player | Reported status | Decision-making impact · qualitative inference |
|---|---|---|
| BUF · Ed Oliver | Out · knee | Removes a defensive-line option against Williams and Stafford. Canonical defensive sample: three games, four QB hits, 0.5 sack; no numerical points adjustment. |
| BUF · Josh / Joshua Palmer | Out · shoulder | Receiver depth loss. Canonical “Josh Palmer” sample: three games, nine targets, 80 yards, two receiving TDs. Injury report uses Joshua Palmer. |
| BUF · DJ Moore | Questionable · shoulder; limited Saturday | Three-game receiving sample: 20 targets, 178 yards. An absence would increase uncertainty over target distribution. |
| BUF · Dee Alford / Landon Jackson | Questionable · ankle / knee | Coverage and defensive-front depth remain uncertain. |
| BUF · Christian Benford | No designation; expected to start after missing Week 4 | His return may improve the secondary relative to the most recent lineup. Maxwell Hairston is expected opposite him; Skyler Bell is expected active for his debut. |
| LAR · Aaron Donald / Jaylen Watson | No designation · back / shoulder | Expected returns after missing Philadelphia improve front and coverage options against Allen. Donald's accepted defensive sample is only two games. |
| LAR · Colby Parkinson | No designation · knee/shoulder | Expected return adds a tight-end option; canonical receiving sample is three games, five targets. |
| LAR · Quentin Lake | Questionable · hip; limited Saturday | Safety availability matters to containing Buffalo's explosive plays; do not assume clearance. |
Sources: Bills final injury report [https://www.buffalobills.com/news/buffalo-bills-injury-report-vs-rams-week-5] · Rams final injury report [https://www.therams.com/news/rams-injury-report-week-5-2026-vs-bills-aaron-donald-colby-parkinson-and-jaylen-watson-good-to-go-for-monday-night-football]. These reports describe expected availability, not confirmed game participation.
Special teams · separate phase
Buffalo has the early edge
A 3.92 EPA-per-game difference favors Buffalo across four covered games. It can matter to a close spread, but is not a 3.92-point adjustment to this game's score.
Bass is 6-for-6 on field goals and 13-for-14 on extra points; Mevis is 5-for-7 and 10-for-10. Small kicking samples are not reliable future make probabilities.
2026, four games each; metrics as of October 4, 17:00 UTC. Special-teams ranks unavailable; expected field-goal value not modeled. Return/field-position and officiating effects are not quantified.
Venue, weather, schedule and rules
Covered field · outdoor airflow
NWS nearby Hawthorne point, 33.93°N / 118.33°W: October 12 at 5 p.m. PDT, 15 minutes before kickoff. Issued October 11, 11:26 a.m. PDT; observed October 12 during 08:48–08:49 UTC. Forecast, not stadium observation. NWS hourly table [https://forecast.weather.gov/MapClick.php?CityName=Inglewood&FcstType=digital&lat=33.9345&lg=ep&lon=-118.332&site=LOX&state=CA]
The database labels the roof “dome” and the surface “matrixturf”; SoFi's architect [https://www.hksinc.com/our-news/articles/inside-sofi-stadiums-innovative-design/] describes an open-sided canopy that permits airflow. Field-level wind and temperature are unknown. The canopy reduces direct rain exposure; no numerical weather adjustment is justified.
Canonical context: equal rest, Los Angeles home, Buffalo traveling west; no measured travel penalty. NFL 2026 profile: four 15-minute quarters, continuous first-down clock, ten-minute regular-season overtime with guaranteed possession and ties allowed. Bookmaker overtime settlement remains unknown.
Accepted prospective forecasts
Available predictions · substantial disagreement
14 forecasts · 34 pregame snapshots · preparation current. Nine models support margin; three support total. All forecast rows use completed prior periods only: 64 league games, training cutoff October 9, 00:15 UTC; latest training kickoff October 6, 00:15 UTC. Calculated October 12, 07:01:19 UTC. Target outcomes and postgame snapshots remain null.
Margin evidence near Rams -3.5
Those models rank first and second on margin MAE, but fall on opposite sides of 3.5. Across all supported margins, predictions range from Rams +1.36 to +10.83. Rams win estimates span 47.33%–84.11%; this is model disagreement, not a confidence interval.
Total forecasts on a common 40–75 scale
All 14 accepted model outputs and declared purposes
| Model / purpose | Rams win estimate | Home margin | Total |
|---|---|---|---|
| Margin Elopredictive | 54.9% | +1.36 | — |
| AMF Powercomposite | 65.6% | +4.20 | 69.38 |
| Predictive SRSpredictive | 65.4% | +4.15 | 46.22 |
| Drive Powerpredictive | 63.0% | +3.46 | — |
| Opponent-adjusted EPA Powerpredictive | 84.1% | +10.83 | — |
| Simple Rating Systemdescriptive | 68.3% | +4.98 | — |
| Location-adjusted SRSdescriptive | 58.7% | +2.28 | — |
| Offense / Defense SRSdescriptive | 61.2% | +2.97 | 67.32 |
| Massey Margin Ratingpredictive | 66.3% | +4.39 | — |
| Football Elopredictive | 56.2% | — | — |
| Glicko-2predictive | 54.4% | — | — |
| Bradley-Terrypredictive | 68.7% | — | — |
| Keener Graph Ratingpredictive | 56.0% | — | — |
| Pythagorean Expectationpredictive | 47.3% | — | — |
Positive margin means Rams by that amount. — means the model does not support that output, not a missing game forecast. Descriptive systems retain their descriptive purpose; forecasts are model estimates, not automatically calibrated betting probabilities.
Benchmark evidence limits confidence
| 2026 NFL benchmark | Brier score ↓ | Margin MAE ↓ | Total MAE ↓ |
|---|---|---|---|
| Margin Elo | 0.238 | 10.11 rank 1 | — |
| AMF Power | 0.253 | 11.32 rank 2 | 19.90 |
| Predictive SRS | 0.303 | 16.61 rank 5 | 12.40 |
| Simple Rating System | 0.280 | 14.66 rank 4 | — |
| Offense / Defense SRS | 0.314 | 19.79 rank 9 | 19.64 |
Each displayed benchmark: 78 probability and margin observations; three total-supporting systems also have 78 total observations. Traditional SRS and Margin Elo have zero total observations. League context, non-provisional; calculated October 12, 07:01:19 UTC. Market margin MAE is 12.60 over 78 observations. Formula versions match the named v1 systems; Predictive SRS uses predictive-srs-v1 and traditional SRS uses traditional-srs-v1. This evaluation sample differs from the 64-game training window for this matchup.
Probability boundary: no declared, validated distribution for final margins or totals is supplied. Mean forecasts, MAE/RMSE, descriptive uncertainty and win estimates cannot establish spread-cover or over/under probability. Numerical cover/total probabilities and expected value are therefore withheld.
Value at the supplied prices only
What must happen for each market to work
Rams -180 / Bills +150
Rams case: defensive suppression, returning personnel and the majority of accepted win estimates support an outright direction. AMF's 65.63% estimate is only 1.34 percentage points above the 64.29% price threshold.
Bills case: the latest SRS, offensive efficiency and special teams support resistance. Margin Elo implies 45.13% Buffalo wins versus the 40.00% price threshold; AMF implies only 34.37%.
Value judgment: neither side's apparent edge survives model choice and calibration concerns. No consensus fair price or quantified EV is asserted.
Preferred direction: Bills +3.5 (-115)
Cover case: Buffalo wins or loses by no more than three on the settled final margin. Its explosive run/QB attack, favorable special-teams sample and Margin Elo's Rams +1.36 mean support that path.
Opposing Rams -3.5 (-105) case: a win by four or more, driven by containing Buffalo's explosives and exploiting Oliver's absence. AMF's Rams +4.20 and Predictive SRS's +4.15 means support that direction.
Uncertainty / value: AMF is just 0.70 points above 3.5 while its margin MAE is 11.32; that is error context, not a probability calculation. Bills -115 needs 53.49% cover, Rams -105 needs 51.22%; neither is established. Pending receiver/secondary availability further weakens conviction.
Preferred direction: Under 54.5 (-105)
Under case: Los Angeles suppresses Buffalo's explosives, receiver absences limit options, and the Rams continue to convert yardage into relatively few points. Predictive SRS's 46.22 mean supports this path and has the lowest total MAE among the three total models: 12.40.
Over 54.5 (-115) case: Nacua's return helps Los Angeles exploit Buffalo's defense while Allen/Cook retain their big-play production. AMF's 69.38 and descriptive Offense / Defense SRS's 67.32 means support it.
Uncertainty / value: the under needs 51.22% and over 53.49%. Model means and weather context do not establish either chance; total error and the 23.17-point model split warrant PASS.