SP+
–
‹ Week 9ProjectionsMetrics explained
Flattered and undersold
Flattered · the score ran ahead of the play by more than 7 points
- UCLA at Minnesota, 34–20 W+14 vs +3 implied · +11The scoreboard flattered UCLA: the +14 margin ran 11 points ahead of the +3 implied by per-play quality (PPA (expected points added per play) +0.17 vs +0.10, success rate 42% vs 33%) (Minnesota turned it over once).
- SMU vs Virginia Tech, 51–28 W+23 vs +12 implied · +11The scoreboard flattered SMU: the +23 margin ran 11 points ahead of the +12 implied by per-play quality (PPA (expected points added per play) +0.42 vs +0.25, success rate 53% vs 48%) (Virginia Tech turned it over twice).
- Miami at Notre Dame, 24–31 L-7 vs -16 implied · +9The scoreboard flattered Miami: the -7 margin ran 9 points ahead of the -16 implied by per-play quality (PPA (expected points added per play) +0.04 vs +0.27, though Miami was more consistent: success rate 43% vs 38%).
- Wisconsin vs Rutgers, 31–10 W+21 vs +12 implied · +9The scoreboard flattered Wisconsin: the +21 margin ran 9 points ahead of the +12 implied by per-play quality (PPA (expected points added per play) +0.18 vs -0.05, success rate 39% vs 38%).
- South Carolina vs Texas A&M, 17–9 W+8 vs +0 implied · +8The scoreboard flattered South Carolina: the +8 margin ran 8 points ahead of the -0 implied by per-play quality (PPA (expected points added per play) +0.13 vs +0.09, though Texas A&M was more consistent: success rate 36% vs 42%).
Undersold · the score ran behind the play by more than 7 points
- Missouri vs Texas, 20–34 L-14 vs -5 implied · -9The scoreboard undersold Missouri: the -14 margin ran 9 points behind the -5 implied by per-play quality (PPA (expected points added per play) +0.04 vs +0.10, success rate 37% vs 40%) (Missouri turned it over once).
- Texas Tech vs West Virginia, 53–48 W+5 vs +13 implied · -8The scoreboard undersold Texas Tech: the +5 margin ran 8 points behind the +13 implied by per-play quality (PPA (expected points added per play) +0.41 vs +0.28, success rate 43% vs 43%).
- Louisville at Georgia Tech, 44–17 W+27 vs +34 implied · -7The scoreboard undersold Louisville: the +27 margin ran 7 points behind the +34 implied by per-play quality (PPA (expected points added per play) +0.37 vs -0.14, success rate 54% vs 32%).
Closest projection
Over the 25 games this week where SP+, Elo and the market all existed before kickoff.
| model | avg miss | closest |
|---|---|---|
| SP+ | 13.6 | 2 of 25 |
| Elo | 12.6 | 14 of 25 |
| Market | 12.6 | 9 of 25 |
Market only (no SP+ or Elo rating for an FCS opponent): 1 game, avg miss 29.5; not in the table.
Our model experimental · 26 eligible games · vs SP+: 14.1 against 13.6, closest 12 / 13 · vs Elo: 14.1 against 12.6, closest 7 / 18 · vs Market: 14.1 against 12.6, closest 8 / 18 · all four: our model closest 3 of 25
Our call · 25 games where SP+, Elo and Our model all existed before kickoff, and each model on those same games, the market among them:
| model | games | avg miss | named the winner |
|---|---|---|---|
| Our call | 25 | 13.2 | 18 of 25 |
| SP+ | 25 | 13.6 | 18 of 25 |
| Elo | 25 | 12.6 | 19 of 25 |
| Market | 25 | 12.6 | 18 of 25 |
| Our model | 25 | 14.1 | 17 of 25 |
Our call against the market, head to head: Our call was at least as close in 11 of 25, the market in 14 (a tie counts for both).
Who's been right: every model
The scorecard: every model by week, with Our call's biggest misses and best calls ›
Each model is graded on every game it predicted before kickoff, from the first game of the season. The market is graded like every model.
| model | named the winner | avg miss | from |
|---|---|---|---|
| Market | 195 of 245 · 80% | 12.1 | Wk 1 |
| Elo | 167 of 222 · 75% | 12.9 | Wk 1 |
| SP+ | 141 of 192 · 73% | 11.6 | Wk 2 |
| Our model | 115 of 157 · 73% | 12.8 | Wk 4 |
| ESPN FPI | 52 of 71 · 73% | 12.7 | Wk 8 |
| Sagarin | 52 of 71 · 73% | 12.3 | Wk 8 |
| The Margin | 52 of 71 · 73% | 12.3 | Wk 8 |
| SRS | 52 of 72 · 72% | 12.7 | Wk 8 |
| Our call | 108 of 153 · 71% | 12.3 | Wk 4 |
| Sonny Moore | 49 of 71 · 69% | 11.9 | Wk 8 |
| Record | 140 of 204 · 69% | – | Wk 2 |
Market for Weeks 1, 2, 4, 6 (55 games) and Elo for Week 1 (30 games) come from their published pre-game numbers (CollegeFootballData), read after the fact: reconstructed, never a number published after kickoff.
Record (ours) names a winner only: the better record before the week, the home team on equal records. ESPN FPI and SRS come from CollegeFootballData's ratings; Sagarin (his PREDICTOR rating), Sonny Moore and The Margin from their own pages, credited by name on our models page. A game counts only when the model's number was saved before its kickoff.
Season to date through Week 10. The table above is the head to head: the models on the same games this week.
Upsets
Ranked teams that lost to an unranked or lower-ranked opponent, by the AP Top 25 in effect for week 10, with each model's pregame margin from the ranked side, the market among them; ✓ when a model had the winner, ✗ when it did not.
- #1 Miami lost at #4 Notre DameNotre Dame 31, Miami 24Elo ✓-5.7Market ✓-3.5Our model ✓-11.1
-
SP+–Elo ✓-10.3Market ✓-5.5Our model ✓-10.0
- #11 Ohio State lost vs OregonOregon 41, Ohio State 28SP+–Elo ✗+11.4Market ✗+10.5Our model ✗+9.4
-
SP+–Elo ✗+0.6Market ✗+5.5Our model ✗+3.9
Called it: SP+ 0 of 0 · Elo 2 of 4 · Market 2 of 4 · Our model 2 of 4.
Biggest swings
The plays that moved win probability most, across the 26 games with win-probability data, from the home side.
- Vanderbilt at Mississippi StateMississippi State 55% → 9%Q4 0:30, 1st & 10: (00:30) Kneel down by Vanderbilt at State46 for loss of 1 yard · final Mississippi State 23, Vanderbilt 25
- Vanderbilt at Mississippi StateMississippi State 10% → 55%Q4 0:37, 2nd & 2: (00:37) No Huddle-Shotgun #17 A.Swann pass complete deep right to #13 S.Magee caught at Vandy00, for 30 yards to the Vandy00 TOUCHDOWN, clock 00:32, 1ST DOWN #17 A.Swann pass attempt failed · final Mississippi State 23, Vanderbilt 25
- UCLA at MinnesotaMinnesota 62% → 18%Q4 12:48, 1st & 10: (12:48) No Huddle-Shotgun #14 T.Dieffenbach pass complete deep middle to #1 A.Mizell caught at MINN40, for 75 yards to the MINN00 TOUCHDOWN, clock 12:41, 1ST DOWN #15 M.Bhaghani kick attempt good (H: #45 C.Barry, LS: #48 G.Cereghino) · final Minnesota 20, UCLA 34
Regression watch, checked
Last week's teams whose score ran more than 7 points from the play, and what they did this week against their season-to-date form (average margin before the game, home field removed). The 2022–25 study found flattered teams fell -3.0 short of form the next game and undersold teams beat it by +1.7.
- Georgia · flattered +20 vs Floridaat Ole Miss 42–7 W · +12.3 vs form (+17.7 vs the field)
- Iowa · flattered +15 vs Wisconsinat Northwestern 24–10 W · +1.4 vs form (+6.7 vs the field)
- Oregon · flattered +14 vs Northwesternat Ohio State 41–28 W · +4.2 vs form (+9.6 vs the field)
- Utah · flattered +14 vs Cincinnativs BYU 28–20 W · -16.9 vs form (-11.5 vs the field)
- USC · flattered +14 vs Ohio Statedid not play
- Ole Miss · flattered +12 vs Auburnvs Georgia 7–42 L · -50.3 vs form (-44.9 vs the field)
- BYU · flattered +12 vs Arizona Stateat Utah 20–28 L · -18.5 vs form (-13.1 vs the field)
- Notre Dame · flattered +10 vs Navyvs Miami 31–24 W · -22.5 vs form (-17.1 vs the field)
- Pittsburgh · flattered +10 vs Georgia Techdid not play
- Florida · undersold -20 vs Georgiavs Oklahoma 7–23 L · -25.7 vs form (-20.3 vs the field)
- Wisconsin · undersold -15 vs Iowavs Rutgers 31–10 W · +12.3 vs form (+17.7 vs the field)
- Northwestern · undersold -14 vs Oregonvs Iowa 10–24 L · -30.8 vs form (-25.4 vs the field)
- Ohio State · undersold -14 vs USCvs Oregon 28–41 L · -36.6 vs form (-31.3 vs the field)
- Penn State · undersold -13 vs Purdueat Washington 28–25 W · -4.2 vs form (+1.1 vs the field)
- Auburn · undersold -12 vs Ole Missvs Arkansas 26–23 W · +3.4 vs form (+8.8 vs the field)
- Florida State · undersold -11 vs Clemsonat Boston College 31–24 W · +8.4 vs form (+13.7 vs the field)
- Michigan · undersold -10 vs Rutgersvs Michigan State 6–15 L · -15.1 vs form (-9.8 vs the field)
- SMU · undersold -9 vs Syracusevs Virginia Tech 51–28 W · +8.1 vs form (+13.5 vs the field)
- Louisville · undersold -9 vs Stanfordat Georgia Tech 44–17 W · +17.1 vs form (+22.4 vs the field)
- Indiana · undersold -8 vs Minnesotadid not play
Baseline: all 39 page teams that played ran -5.4 vs form this week. The baseline matters because early-season form comes mostly from easier non-conference games, so the whole field tends to fall short of it once conference play starts; a flagged group is read against the field, not against zero.
Flattered: -12.9 vs form over 7, -7.5 against the baseline (-3.0 in the study) · Undersold: -6.3 vs form over 10, -0.9 against the baseline (+1.7 in the study). One week is a few teams and proves nothing either way; the point is that the check is visible every week.
How these are computed
Verdicts are the team pages' own: the score against the margin the plays earned, recomputed here from the box score and drives, so any past week rebuilds the same way. The closest-projection table is the accuracy report's block for the week; a game counts only if one of its teams had a page before kickoff, and our model is kept on its own line. SP+ is graded only in weeks where its own rating was posted before kickoff; an earlier snapshot shown before a game is never scored. Upsets use the AP Top 25 in effect for the week's games; every ranked team has a page, so no upset is missed. Pregame margins are the same ones the season record scores. Swings are the win-probability moves in the play-by-play. The regression check uses the study's own definition of form. Nothing on this page is written by AI, and a section with nothing to show is left out.