Sun, Sep 12 2021
·
Week 2
·
🏟 Michigan Stadium
Ann Arbor, MI
·
Turf
·
107,601 cap
Washington✈ 1,900 mi+3 hr TZ
Matchup Prediction
Toss-up — no clear edge
Neither metric shows a meaningful pre-game edge in this matchup.
Momentum Control
58.4%
—
Lean
Game Control
76%
Michigan wins
Strong
Vegas Spread
Michigan -6.5
O/U 47.5
teamrankings
Advanced Stats
All 4 factors agree → Michigan
· 83.1% ATS historically when all four align
↓ See full breakdown
Washington 2021 Schedule
Washington's 2021 Schedule
| Date | Matchup | Spread | Total | Result | O/U | Cover |
|---|---|---|---|---|---|---|
| Sat 9/4 | Washington vs Montana | -22.5L7–13 | 54.0 | L7–13 | U | N |
| Sat 9/11 | Washington at Michigan | +6.5L10–31 | 47.5 | L10–31 | U | N |
| Sat 9/18 | Washington vs Arkansas State | -17.5W52–3 | 58.5 | W52–3 | U | Y |
| Sat 9/25 | Washington vs California | -7.5W31–24 | 47.5 | W31–24 | O | N |
| Sat 10/2 | Washington at Oregon State | +2.5L24–27 | 57.5 | L24–27 | U | N |
| — Bye Week — | ||||||
| Sat 10/16 | Washington vs UCLA | -1.5L17–24 | 55.5 | L17–24 | U | N |
| Fri 10/22 | Washington at Arizona | -17.5W21–16 | 45.5 | W21–16 | U | N |
| Sat 10/30 | Washington at Stanford | +2.5W20–13 | 45.5 | W20–13 | U | Y |
| Sat 11/6 | Washington vs Oregon | +7.0L16–26 | 48.0 | L16–26 | U | N |
| Sat 11/13 | Washington vs Arizona State | +6.0L30–35 | 45.5 | L30–35 | O | Y |
| Sat 11/20 | Washington at Colorado | -6.5L17–20 | 43.0 | L17–20 | U | N |
| Fri 11/26 | Washington vs Washington State | -1.0L13–40 | 45.0 | L13–40 | O | N |
Michigan 2021 Schedule
Michigan's 2021 Schedule
| Date | Matchup | Spread | Total | Result | O/U | Cover |
|---|---|---|---|---|---|---|
| Sat 9/4 | Michigan vs Western Michigan | -16.5W47–14 | 65.5 | W47–14 | U | Y |
| Sat 9/11 | Michigan vs Washington | -6.5W31–10 | 47.5 | W31–10 | U | Y |
| Sat 9/18 | Michigan vs Northern Illinois | -27.5W63–10 | 54.5 | W63–10 | O | Y |
| Sat 9/25 | Michigan vs Rutgers | -20.0W20–13 | 50.0 | W20–13 | U | N |
| Sat 10/2 | Michigan at Wisconsin | +2.0W38–17 | 43.5 | W38–17 | O | Y |
| Sat 10/9 | Michigan at Nebraska | -2.5W32–29 | 50.5 | W32–29 | O | Y |
| — Bye Week — | ||||||
| Sat 10/23 | Michigan vs Northwestern | -23.5W33–7 | 51.5 | W33–7 | U | Y |
| Sat 10/30 | Michigan at Michigan State | -4.0L33–37 | 50.5 | L33–37 | O | N |
| Sat 11/6 | Michigan vs Indiana | -20.5W29–7 | 51.0 | W29–7 | U | Y |
| Sat 11/13 | Michigan at Penn State | -2.5W21–17 | 48.0 | W21–17 | U | Y |
| Sat 11/20 | Michigan at Maryland | -16.0W59–18 | 58.5 | W59–18 | O | Y |
| Sat 11/27 | Michigan vs Ohio State | +6.5W42–27 | 63.5 | W42–27 | O | Y |
| Sat 12/4 | Michigan vs Iowa | -12.0W42–3 | 43.5 | W42–3 | O | Y |
| Fri 12/31 | Michigan vs Georgia | +7.5L11–34 | 47.0 | L11–34 | U | N |
Advanced Stats
Advanced Analytics Matchup
Matchup-adjusted (offense vs opponent defense) ·
2021 season
Agreement Signals — When All Metrics Agree
Elite · 83.1% ATS
PPA + PPO + SR + Havoc
All 4 Agree
→ Michigan
Elite · 82.4% ATS
PPA + PPO + Havoc
3 Agree
→ Michigan
Elite · 73.9% ATS
PPA + Success Rate
Both Agree
→ Michigan
Individual Factors — Ranked by Predictive Strength
PPA Overall
Points added per play · Elite predictor
PPA Passing
Pass efficiency edge · Strong predictor
Havoc Total
Def. disruption rate · Strong predictor
TFLs, sacks, PBUs, forced fumbles — higher is better
Points Per Opp
Drive-finishing edge · Strong predictor
Success Rate
Play consistency edge · Solid predictor
Field Position
Avg start (lower=better) · Solid predictor
Avg yards from own endzone to average start — lower is better · longer bar = better field position
Advanced stats sourced from CFBD · 2021 season ·
Edges are matchup-adjusted (offense vs opponent defense)
Power Ratings
Team Power Ratings
Overall · Offense · Defense ratings · Updated as season progresses
Power ratings updated throughout the season as results accumulate
Momentum Control (CSS)
Consecutive Scoring Sequences
Who builds scoring momentum?
Washington Edge
Washington +0.00
CSS Edge (season-to-date)
Teams with this edge win 58.4% of games historically
Based on 1 game this season
Game Control (GC)
Win Probability Dominance
Who controls games start to finish?
Michigan Edge
Michigan +23.3
GC Edge (season-to-date)
Teams with this edge win 76% of games historically
Based on 1 game this season
Spread Context
ATS Historical Context
Based on 2021–2025 backtest · FBS vs FBS · Regular season
Both metrics agree on Michigan with a large edge. Historically, dominant teams like this are fully priced into the spread — the agreed-upon team covers just 50.2% of the time. The metrics predict game control better than they beat the number.
ATS data is informational only. Past cover rates do not guarantee future results.
Coaching Matchup
Washington
Jimmy Lake #1
4–3 (57%)
· Yr 2 at school
OC
John Donovan
Yr 1
#1
DC
Ikaika Malloe
Yr 1
#1
Michigan
Jim Harbaugh #1
52–22 (70%)
· Yr 7 at school
OC
Josh Gattis
Yr 1
#1
DC
Mike Macdonald
Yr 1
#1
About these metrics
Advanced Stats shows matchup-adjusted factor edges (offense vs opponent defense). Combination signals — when PPA, PPO, Success Rate, and Havoc all point the same direction — have historically predicted the SU winner in 95–97% of games and the ATS winner in 82–83% of games (2021–2025, FBS vs FBS, regular season).
Impact: Advanced Stats are the best performance based metric used to predict the outcome of games. ✓
Momentum Control (CSS) measures consecutive scoring sequences — when a team scores, holds the opponent scoreless, then scores again. Teams entering a game with a CSS edge of +1.0 or more have won 71–78% of games historically (2021–2025, FBS vs FBS).
Impact: Momentum Control is a great measure for predicting game outcome but NOT an ATS advantage, data shows this is already considered when lines are set. ✗
Game Control (GC) measures win probability dominance — how thoroughly a team controlled the game from start to finish. Teams with a GC edge of +12 or more have won 67–76% of games historically. When both metrics agree, combined confidence is higher. When they split, treat as a lean at best.
Impact: Game Control is another great measure for predicting game outcome but NOT an ATS advantage, data shows this is already considered when lines are set. ✗
Power Ratings are a custom-built composite of a Teams Talent, Experience & Production, Coaching & Performance Metrics. These are updated constantly with roster changes, performance once the games start for the 2026 season, injuries the team is dealing with and scheduling situations.
Impact: There are a wide range of power ratings available, we think ours is the best, you can decide for yourself ✓
Advanced Stats shows matchup-adjusted factor edges (offense vs opponent defense). Combination signals — when PPA, PPO, Success Rate, and Havoc all point the same direction — have historically predicted the SU winner in 95–97% of games and the ATS winner in 82–83% of games (2021–2025, FBS vs FBS, regular season).
Impact: Advanced Stats are the best performance based metric used to predict the outcome of games. ✓
Momentum Control (CSS) measures consecutive scoring sequences — when a team scores, holds the opponent scoreless, then scores again. Teams entering a game with a CSS edge of +1.0 or more have won 71–78% of games historically (2021–2025, FBS vs FBS).
Impact: Momentum Control is a great measure for predicting game outcome but NOT an ATS advantage, data shows this is already considered when lines are set. ✗
Game Control (GC) measures win probability dominance — how thoroughly a team controlled the game from start to finish. Teams with a GC edge of +12 or more have won 67–76% of games historically. When both metrics agree, combined confidence is higher. When they split, treat as a lean at best.
Impact: Game Control is another great measure for predicting game outcome but NOT an ATS advantage, data shows this is already considered when lines are set. ✗
Power Ratings are a custom-built composite of a Teams Talent, Experience & Production, Coaching & Performance Metrics. These are updated constantly with roster changes, performance once the games start for the 2026 season, injuries the team is dealing with and scheduling situations.
Impact: There are a wide range of power ratings available, we think ours is the best, you can decide for yourself ✓