Week 10 · 2026
Buffalo at Miami (OH)
Visual game report
Available after a final score is recorded. The pregame forecast stays separate from postgame results.
Player production & success
CFBD regular season · through week 9 · observed 2026-11-06 UTC. Retrospective prior-game summary; these observations were not verified available before kickoff.
Most active players in up to 3 recent completed team games. Passing and rushing are separate credited populations; rates weight successes by plays. Garbage time included; minimum credited plays 0. Partial provider coverage; no forecast inputs.
Buffalo
Window: 2026-10-24 (reported) · 2026-10-17 (reported) · 2026-10-10 (reported). Cutoff: 2026-11-04T01:00:00+00:00.
| Player | Passing | Rushing |
|---|---|---|
| Jason Wright QB 3 games | 35.9% 33 / 92 plays | 50.0% 9 / 18 plays |
| Messiah Burch RB 3 games | — 0 / 0 plays | 36.4% 28 / 77 plays |
| Nino Hill RB 3 games | — 0 / 0 plays | 52.4% 11 / 21 plays |
| Nino Francavilla OL 1 games | — 0 / 0 plays | 0.0% 0 / 1 plays |
Traditional season production · 0 players
Provider totals through week 9, captured 2026-11-06. Original statistical values; team-specific athlete records remain separate from adjusted efficiency and usage.
Season totals were captured after kickoff or the kickoff is unresolved; they are unavailable for this preview.
Miami (OH)
Window: 2026-10-24 (reported) · 2026-10-17 (reported) · 2026-10-10 (reported). Cutoff: 2026-11-04T01:00:00+00:00.
| Player | Passing | Rushing |
|---|---|---|
| Caleb Heavner QB 3 games | 28.6% 22 / 77 plays | 60.9% 14 / 23 plays |
| Rodney Nelson RB 3 games | — 0 / 0 plays | 41.2% 14 / 34 plays |
| D'Shawntae Jones RB 3 games | — 0 / 0 plays | 41.4% 12 / 29 plays |
| Lazaro Rogers RB 2 games | — 0 / 0 plays | 42.9% 3 / 7 plays |
| Scoop Smith WR 3 games | — 0 / 0 plays | 100.0% 4 / 4 plays |
| Damarion Witten WR 1 games | 0.0% 0 / 1 plays | — 0 / 0 plays |
Traditional season production · 0 players
Provider totals through week 9, captured 2026-11-06. Original statistical values; team-specific athlete records remain separate from adjusted efficiency and usage.
Season totals were captured after kickoff or the kickoff is unresolved; they are unavailable for this preview.
Pregame weather
Decision-time forecast unavailable
No retained forecast matches this scheduled game.
Weather context
Weather record unavailable
Evidence cutoff: . This is an artifact cutoff; verified pregame availability remains unavailable.
No matching weather context has been published for this game.
- Verified pregame availability
- Unavailable
- Measurement units
- Unverified · numeric interpretation withheld
- Forecast issue time
- Unavailable
- Forecast valid time
- Unavailable
Capture timing does not establish whether these provider values describe forecasts or observed conditions.
Latest published card · In-season model
Saved CFBD odds snapshot captured
Saved prices from the latest card, not the weekly freeze. Review the weekly card.
No pick
Miami (OH) -6.5
DraftKings snapshot: Miami (OH) -6.5
Miami (OH) by 5.10
1.40 points toward Buffalo versus the spread. gap below pick minimum The gap is not a cover probability.
Explain this EPA-OLS spread forecast
Baseline: 0.00 Home Margin points. Zero-point accounting baseline; the fitted constant offset is shown separately
Signed contributions in Home Margin points; positive moves toward Miami (OH).
- Fitted offset: +5.20
- Fitted home field: -1.20
- Net EPA rating difference: +1.10
Reconciles to 5.10 points; numerical tolerance 1e-10. Published rounding tolerance 0.0050000001 in output units.
Fit arithmetic, not causal effects or cover probability.
Fit c4bb7cb5f7676bad6ca935c9ad6de236fbdb43b95c02a90800020598436e58af · inputs 79dad43d515c0caab7947eef6cbec17f50b17298ad995537c985ec7f5b5d1eb5 · training 84480cb6cc00c8fa67ed4375e0492a83b5ad7dac628fe1cf824ff35a490cdd51 · forecast 176744f5119f49cec1faad3b33c000194dfb6009fa8450c4edc006ccabf7e602 · through week 9.
What changed?
7 saved cards for this gameNo change in spread, forecast or model gap.
The earlier saved card is not necessarily the frozen weekly selection.
Scroll sideways to compare columns. Row labels stay visible.
| Recorded value | Earlier | Latest | Change |
|---|---|---|---|
| Market spread | -6.5 | -6.5 | 0.0 |
| Model spread | -5.1 | -5.1 | 0.0 |
| Model gap | -1.4 | -1.4 | 0.0 |
Negative spreads favor Miami (OH); positive model gap favors them too.
Capture times & sources
Earlier saved card
Generated
Saved CFBD odds snapshot captured
In-season model · rating source as of unknown
Latest published card
Generated
Saved CFBD odds snapshot captured
In-season model · rating source as of unknown
Model gap = market spread minus model spread.
This compares two saved cards, not every line change. The CFBD market spread is a median, so provider coverage can move it. Capture time may differ from quote time.
Model movement compares published projections. These records do not establish why a projection changed or whether the model code changed.
- Earlier saved card SHA-256
- 09ae362bba85693dce95202484e390b841fbdf36d21ea093fee4dd02276f20a9
- Odds source SHA-256
- 1191efda55b02705391f90dc98ac55459f647f2300778ae80eca48d727f2d3cf
- Latest published card SHA-256
- 1cb82609908ef20168d6821e20ce6db6dd11891a77105bc3ac9b18df06cadb8b
- Odds source SHA-256
- d0dba70812098c221b0c8c478c2238c95459bef99e39ca41d677e1f30597fc3c
Model context & pick method
Pick logic
Card generated . Captured prices are not live quotes.
- Output:
- Home margin and model–spread gap (points)
- Model:
- EPA-OLS
- Training:
- 2026, weeks 5, 6, 7, 8, 9; 285 games. Training snapshot SHA-256 recorded: 84480cb6cc00c8fa67ed4375e0492a83b5ad7dac628fe1cf824ff35a490cdd51. Private linear fit c025ead12920ba4b56114f7cd4fe71d589837d1d707e57540b143a0036e10027; producer verified saved inference without refitting. Raw provider availability is not established
- Features:
- EPA net difference; Feature snapshot SHA-256 recorded: 79dad43d515c0caab7947eef6cbec17f50b17298ad995537c985ec7f5b5d1eb5. Private linear fit c025ead12920ba4b56114f7cd4fe71d589837d1d707e57540b143a0036e10027; producer verified saved inference without refitting. Raw provider availability is not established
- Evaluation:
- EPA-OLS 2025 historical ATS reference. Matching probability calibration unavailable; reference buckets do not validate this card.
Training: 285 games · weeks 5, 6, 7, 8, 9.
Training-only RMSE: 15.77 points. This describes the fitted games, not predictive accuracy or a cover probability.
No pick · gap below pick minimum · gap -1.4 points, minimum 3.0
Win probabilities · separate sources
Miami (OH) win probability by source
- EPA-OLS calibrated win probability
- Unavailable
- Unavailable: no matching calibration for this card’s projection basis
- CFBD market-based probability
- 67%
- Derived from the available point spread, provider snapshot captured Nov 6, 2026, 06:01:15 CST; CFBD source spread: Miami (OH) -6.5
- Stacked ensemble
- 67%
- Stacked ensemble from the model pack
Explain this StackedEnsemble probability
Baseline: 36.75 % home win probability. Observed training row nearest standardized medians, row 1641.
Signed contributions in percentage points; positive moves toward Miami (OH).
- home elo: -11.40 (path difference 8.61)
- away adjusted success: +10.74 (path difference 1.54)
- away elo: +10.21 (path difference 6.53)
- home adjusted rushing epa allowed: +8.64 (path difference 1.00)
- away adjusted passing down success: -6.16 (path difference 2.66)
- away adjusted epa: +6.10 (path difference 1.15)
- home adjusted epa: +5.90 (path difference 4.05)
- home adjusted success: -4.92 (path difference 3.18)
- home adjusted explosiveness: +4.08 (path difference 0.23)
- away adjusted passing down success allowed: -3.97 (path difference 1.01)
- home adjusted pass explosiveness allowed: +3.61 (path difference 3.49)
- home adjusted open field yards allowed: +3.09 (path difference 3.03)
- away adjusted rushing epa allowed: +3.08 (path difference 1.84)
- away adjusted success allowed: +2.91 (path difference 1.44)
- home adjusted passing epa: -2.90 (path difference 1.66)
- home adjusted success allowed: -2.53 (path difference 1.33)
- away adjusted second level yards: +2.46 (path difference 0.29)
- home adjusted second level yards: +2.23 (path difference 1.09)
- home conference: -2.06 (path difference 0.94)
- away conference: -2.05 (path difference 0.90)
- away adjusted second level yards allowed: +1.89 (path difference 3.35)
- home talent: +1.71 (path difference 0.06)
- home adjusted standard down success: -1.57 (path difference 1.50)
- home adjusted epa allowed: -1.47 (path difference 2.01)
- away adjusted explosiveness: +1.39 (path difference 0.69)
- away adjusted passing epa: +1.34 (path difference 1.24)
- week: -1.24 (path difference 0.02)
- away adjusted pass explosiveness: +1.07 (path difference 0.39)
- home adjusted explosiveness allowed: -1.00 (path difference 0.50)
- home adjusted line yards: -0.97 (path difference 0.61)
- home adjusted second level yards allowed: +0.96 (path difference 0.98)
- away adjusted rush explosiveness allowed: +0.88 (path difference 0.14)
- away adjusted line yards allowed: +0.87 (path difference 2.91)
- away talent: +0.78 (path difference 0.91)
- away adjusted rush explosiveness: -0.77 (path difference 1.09)
- away adjusted standard down success: -0.76 (path difference 0.40)
- home adjusted passing down success: +0.72 (path difference 1.07)
- away adjusted epa allowed: -0.71 (path difference 0.02)
- home adjusted passing epa allowed: -0.69 (path difference 0.41)
- away adjusted standard down success allowed: +0.69 (path difference 0.63)
- home adjusted rush explosiveness: +0.66 (path difference 0.10)
- away adjusted open field yards allowed: -0.57 (path difference 1.52)
- away adjusted rushing epa: +0.49 (path difference 0.60)
- away adjusted explosiveness allowed: +0.49 (path difference 0.24)
- home adjusted rushing epa: -0.32 (path difference 0.45)
- away adjusted passing epa allowed: -0.26 (path difference 0.16)
- home adjusted standard down success allowed: -0.21 (path difference 0.54)
- home adjusted rush explosiveness allowed: -0.12 (path difference 0.03)
- away adjusted line yards: -0.07 (path difference 0.02)
- away adjusted pass explosiveness allowed: -0.05 (path difference 0.74)
- away adjusted open field yards: +0.04 (path difference 0.29)
- home adjusted line yards allowed: -0.03 (path difference 0.13)
- home adjusted passing down success allowed: +0.00 (path difference 0.00)
- home adjusted pass explosiveness: -0.00 (path difference 0.00)
- home adjusted open field yards: -0.00 (path difference 0.33)
Reconciles to 66.99 %; numerical tolerance 1e-10. Published rounding tolerance 0.0000500001 in output units.
Reference- and order-dependent allocation; not SHAP, causal effects or spread-pick drivers. Hybrid rows may be unrealistic with correlated inputs. Path difference measures order sensitivity, not uncertainty.
Fit e302459ca56c4f194361bcc0ad952fb75b35a71b9ff0feaf8dede9dc1265ad90 · inputs 63953753616bd94030baac8dd43e37e5182ac26ba24e8a0123c3db18884e2426 · training 0837fbbedb3a684b201340cc993067c34179ef0f7ea0e6cc49da8424378a4dd9 · forecast 1411dfc6b9e7be1ef67e025a6baf50a45dfc105d1a5fb8d3ebe1d5ea2286bead · through week 9.
- Output:
- Home win probability (%), display only
- Model:
- StackedEnsemble
- Training:
- Training SHA-256 0837fbbedb3a684b201340cc993067c34179ef0f7ea0e6cc49da8424378a4dd9
- Features:
- Unavailable — not recorded in this artifact
- Evaluation:
- Matching evaluation identity unavailable; separate from EPA/SP+ spread picks and their ATS record
The spread forecast drives the pick. CFBD converts a market spread to probability. The stacked ensemble is a separate forecast. Neither supplies a missing EPA-OLS probability.
Recent rushing game context unavailable before this matchup.
Team matchup statistics
Offense vs defense
2026 season to date · through Week 9, captured Nov 6, 2026, 06:07:26 CST · raw CFBD PPA, not opponent-adjusted EPA.
| Measure | Miami (OH)Offense | BuffaloDefense | BuffaloOffense | Miami (OH)Defense |
|---|---|---|---|---|
| PPA/play | +0.14 55th | +0.09 42nd | +0.04 21st | +0.11 34th |
| Success rate | 40.6% 31st | 39.3% 63rd | 35.8% 3rd | 42.0% 39th |
| Explosiveness | 1.39 97th | 1.22 53rd | 1.19 31st | 1.27 33rd |
| Standard-down PPA/play | +0.08 49th | +0.11 19th | +0.03 33rd | +0.04 48th |
| Passing-down PPA/play | +0.27 57th | +0.06 81st | +0.06 15th | +0.26 29th |
| Havoc allowed / created | 15.6% 39th | 16.1% 55th | 15.7% 38th | 14.0% 24th |
Bars show percentile among FBS teams with that measure; higher is better for the team shown. Team-level source captured Nov 6, 2026, 06:07:26 CST. CFBD advanced statistics source These descriptive rates do not change the game forecast.
CFBD adjusted preview & kicker PAAR
Provider opponent-adjusted efficiency and season kicking context. These observations do not change our model forecast. Included-game cutoff is not supplied; capture and source update times are not games-through dates. CFBD definitions
Preview captured Nov 3, 2026, 6:18:08 AM CST · assembled Nov 3, 2026, 6:18:08 AM CST
Adjusted preview unavailable: kickoff reached. Saved evidence is retained.
2026 kicker PAAR
Points added above replacement; higher is better. Kicking context, not a complete special-teams rating. Captured Nov 3, 2026, 6:18:06 AM CST.
- Kellan McLaughlin · Miami (OH) · PAAR -3.38 · 14 attempts
- Brady Stidham · Miami (OH) · PAAR 0.52 · 1 attempts
- Oliver Hautanen · Buffalo · PAAR 0.08 · 17 attempts
Player recent form
CFBD raw play PPA and observed workload, 2026. Descriptive context; these values do not change our forecast or establish current roster membership or game availability.
Kickoff reached; no retrospective pregame reconstruction
Captured line movement
Provider quotes observed through CFBD. Chicago times (CST/CDT) show when we received the data, not when a sportsbook changed its price. Changes during coverage gaps are unknown. “Quote confirmed” means a complete CFBD snapshot repeated the saved price; heartbeats do not create price observations. Earlier conservative gap markers remain recorded. Ages update every 30 seconds; evidence over 90 minutes old is highlighted. Observation age measures receipt time, not sportsbook price freshness. Published Nov 6, 2026, 14:39:19 CST · Publication age: calculating…. Line-only updates publish no more than once every three hours; daily updates may publish sooner. Checks for updates every minute while this page is visible.
Loading recorded line history… Open this game’s recorded history.
Line history & team context
Card snapshot lines
Miami (OH) spread · saved card's CFBD odds snapshot. Captured: Nov 6, 2026, 03:00:01 CST · Snapshot age: calculating…. These snapshot quotes were not rechecked now. See Captured line movement above for separately timed provider observations and coverage gaps. CFBD-reported openings and first local observations are separate evidence. Capture times are Chicago times, not sportsbook update times; these are not verified closing lines.
DraftKings
CFBD-reported opening: -6.5 · Snapshot spread: -6.5
First locally observed: -6.5 · captured Nov 1, 2026, 03:00:02 CST · Observation age: calculating…
Observation evidence
Capture SHA-256: eeaa60c32ae950903c6eb357c07f8eda76f9de01d095058b85ce148130691012
Total · CFBD-reported opening: 45.5 · Snapshot total: 45.5
Total · First locally observed: 45.5 · captured Nov 1, 2026, 03:00:02 CST · Observation age: calculating…
Total observation evidence
Capture SHA-256: eeaa60c32ae950903c6eb357c07f8eda76f9de01d095058b85ce148130691012
Home moneyline · Snapshot: -265 · Reported opening: not supplied by CFBD
Home moneyline · First locally observed: -230 · captured Nov 1, 2026, 03:00:02 CST · Observation age: calculating…
Home moneyline observation evidence
Capture SHA-256: eeaa60c32ae950903c6eb357c07f8eda76f9de01d095058b85ce148130691012
Away moneyline · Snapshot: +215 · Reported opening: not supplied by CFBD
Away moneyline · First locally observed: +190 · captured Nov 1, 2026, 03:00:02 CST · Observation age: calculating…
Away moneyline observation evidence
Capture SHA-256: eeaa60c32ae950903c6eb357c07f8eda76f9de01d095058b85ce148130691012
Bovada
CFBD-reported opening: -7 · Snapshot spread: -6.5
First locally observed: -7 · captured Nov 2, 2026, 07:37:21 CST · Observation age: calculating…
Observation evidence
Capture SHA-256: f47e17bb763e93498c8e692f9a5e057fd2d1d524ba9cf3f94dff5a9e9ddc5ccb
Total · CFBD-reported opening: 44 · Snapshot total: 45
Total · First locally observed: 44 · captured Nov 2, 2026, 07:37:21 CST · Observation age: calculating…
Total observation evidence
Capture SHA-256: f47e17bb763e93498c8e692f9a5e057fd2d1d524ba9cf3f94dff5a9e9ddc5ccb
Home moneyline · Snapshot: -260 · Reported opening: not supplied by CFBD
Home moneyline · First locally observed: -260 · captured Nov 2, 2026, 07:37:21 CST · Observation age: calculating…
Home moneyline observation evidence
Capture SHA-256: f47e17bb763e93498c8e692f9a5e057fd2d1d524ba9cf3f94dff5a9e9ddc5ccb
Away moneyline · Snapshot: +215 · Reported opening: not supplied by CFBD
Away moneyline · First locally observed: +215 · captured Nov 2, 2026, 07:37:21 CST · Observation age: calculating…
Away moneyline observation evidence
Capture SHA-256: f47e17bb763e93498c8e692f9a5e057fd2d1d524ba9cf3f94dff5a9e9ddc5ccb
Miami (OH)
Net adjusted EPA/play -0.1 · 9 rated weeks · Mid-American
Buffalo
Net adjusted EPA/play -0.1 · 8 rated weeks · Mid-American
Travel
Buffalo travels 392 miles (shorter than half of road trips, 2017–2025) · no time-zone change.
Distance uses the campuses because this schedule has no venue.
The 2017–2025 study found no overall ATS edge. Distance and time-zone groups cover near a coin flip. The 2,500+ mile group is above break-even but dominated by Hawai'i trips, so it is not a betting rule.
Series history
Miami (OH) leads 17–9 in 26 meetings since 1999.
- 2025Miami (OH) at Buffalo37–20 · Miami (OH)
- 2023Buffalo at Miami (OH)10–23 · Miami (OH)
- 2022Miami (OH) at Buffalo20–24 · Buffalo
- 2021Buffalo at Miami (OH)18–45 · Miami (OH)
- 2020Miami (OH) at Buffalo10–42 · Buffalo
- 2019Buffalo at Miami (OH)20–34 · Miami (OH)
CFBD matchup data Last 6 meetings shown. N marks a neutral site. History is not a projection input.
Team style charts
Model output, not betting advice. Historical hit rates are small-sample backtests; several edge buckets lose money at -110. Flat units only.