Week 7 in review · CFB · The Sports Wire

Flattered and undersold

Flattered · the score ran ahead of the play by more than 7 points

BYU vs Notre Dame, 28–24 W+4 vs -8 implied · +12

The scoreboard flattered BYU: the +4 margin ran 12 points ahead of the -8 implied by per-play quality (PPA (expected points added per play) +0.16 vs +0.24, though BYU was more consistent: success rate 46% vs 39%) (Notre Dame turned it over once).
Florida State at Miami, 35–43 L-8 vs -17 implied · +9

The scoreboard flattered Florida State: the -8 margin ran 9 points ahead of the -17 implied by per-play quality (PPA (expected points added per play) +0.06 vs +0.26, success rate 35% vs 44%) (Miami turned it over once).
Utah at Colorado, 27–24 W+3 vs -4 implied · +7

The scoreboard flattered Utah: the +3 margin ran 7 points ahead of the -4 implied by per-play quality (PPA (expected points added per play) +0.01 vs +0.10, success rate 40% vs 36%) (Colorado turned it over twice).

Undersold · the score ran behind the play by more than 7 points

Wake Forest at California, 41–7 W+34 vs +48 implied · -14

The scoreboard undersold Wake Forest: the +34 margin ran 14 points behind the +48 implied by per-play quality (PPA (expected points added per play) +0.52 vs -0.20, success rate 50% vs 33%).
Louisville at Syracuse, 34–26 W+8 vs +21 implied · -13

The scoreboard undersold Louisville: the +8 margin ran 13 points behind the +21 implied by per-play quality (PPA (expected points added per play) +0.19 vs -0.11, success rate 43% vs 43%) (Louisville turned it over twice).
Washington at Purdue, 38–26 W+12 vs +25 implied · -13

The scoreboard undersold Washington: the +12 margin ran 13 points behind the +25 implied by per-play quality (PPA (expected points added per play) +0.25 vs -0.14, success rate 52% vs 33%) (Washington turned it over 4 times).
Northwestern at Michigan State, 24–21 W+3 vs +16 implied · -13

The scoreboard undersold Northwestern: the +3 margin ran 13 points behind the +16 implied by per-play quality (PPA (expected points added per play) +0.23 vs +0.02, success rate 45% vs 44%).
Florida at Texas, 27–24 W+3 vs +14 implied · -11

The scoreboard undersold Florida: the +3 margin ran 11 points behind the +14 implied by per-play quality (PPA (expected points added per play) +0.09 vs -0.13, though Texas was more consistent: success rate 34% vs 41%).

Closest projection

Over the 21 games this week where SP+, Elo and the line all existed before kickoff, the line as a benchmark (its miss only, never closest).

model	avg miss	closest
SP+	12.3	10 of 21
Elo	12.4	11 of 21
Line	12.3	—

Our model experimental · 22 eligible games · vs SP+: 12.3 against 12.3, closest 11 / 10 · vs Elo: 12.3 against 12.4, closest 10 / 11 · vs Line: 12.3 against 12.3, closest 10 / 11 · all four: our model closest 9 of 21

Our call · 21 games where SP+, Elo and Our model all existed before kickoff, each model on those same games, and the line for comparison (its miss only: a benchmark is never counted as naming the winner):

model	games	avg miss	named the winner
Our call	21	12.2	15 of 21
SP+	21	12.3	15 of 21
Elo	21	12.4	13 of 21
Line (benchmark)	21	12.3	—
Our model	21	12.3	14 of 21

Against its benchmark, the line: Our call was at least as close in 11 of 21, the line in 10 (a tie counts for both).

Upsets

Ranked teams that lost to an unranked or lower-ranked opponent, by the AP Top 25 in effect for week 7, with each model's pregame margin from the ranked side and the line beside them for comparison only; ✓ when a model had the winner, ✗ when it did not.

#1 Georgia lost vs AuburnAuburn 27, Georgia 24

SP+ ✗
+19.8

Elo ✗
+17.9

Line
+18.5

Our model ✗
+17.7
#2 Notre Dame lost at #8 BYUBYU 28, Notre Dame 24

SP+ ✗
+4.0

Elo ✗
+12.4

Line
+12.5

Our model ✗
+11.0
#6 Alabama lost at #14 TennesseeTennessee 26, Alabama 10

SP+ ✗
+3.8

Elo ✗
+6.1

Line
+6.5

Our model ✗
+10.0
#7 Texas lost vs FloridaFlorida 27, Texas 24

SP+ ✗
+15.9

Elo ✗
+14.3

Line
+10.0

Our model ✗
+8.7
#16 Mississippi State lost at #22 LSULSU 31, Mississippi State 0

SP+ ✓
-11.6

Elo ✓
-12.4

Line
-11.5

Our model ✓
-10.3
#18 Oklahoma State lost at #20 HoustonHouston 35, Oklahoma State 32

SP+ ✓
-4.4

Elo ✓
-12.1

Line
-3.5

Our model ✓
-15.7

Called it: SP+ 2 of 6 · Elo 2 of 6 · Our model 2 of 6.

Biggest swings

The plays that moved win probability most, across the 22 games with win-probability data, from the home side.

Nebraska at OregonOregon 40% → 92%

Q4 0:28, 2nd & 4: (00:28) Shotgun #8 D.Raiola pass complete short middle to #9 J.Johnson caught at NEB00, for 4 yards to the NEB00 TOUCHDOWN, clock 00:28 #83 H.Lewis pass attempt Successful · final Oregon 17, Nebraska 14
Florida State at MiamiMiami 93% → 47%

Q4 2:46, 1st & 10: (02:46) No Huddle-Shotgun #14 A.Daniels pass complete deep left to #0 D.Robinson caught at UM50, for 75 yards to the UM00 TOUCHDOWN, clock 02:39, 1ST DOWN #39 G.Panikowski kick attempt good (H: #87 C.Jula, LS: #48 C.Bowers) · final Miami 43, Florida State 35
Northwestern at Michigan StateMichigan State 82% → 38%

Q1, 1st & 1: Shotgun #0 A.Chiles rush middle for 1 yard gain to the MSU00 TOUCHDOWN, clock 07:24 #86 A.Honig pass attempt Successful · final Michigan State 21, Northwestern 24

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.

Oregon · flattered +21 vs UCLAvs Nebraska 17–14 W · -18.0 vs form (-5.1 vs the field)
Penn State · flattered +17 vs USCat Michigan 18–21 L · -11.7 vs form (+1.2 vs the field)
Missouri · flattered +15 vs Texas A&Mat Ole Miss 3–31 L · -42.8 vs form (-29.9 vs the field)
Oklahoma State · flattered +15 vs UCFat Houston 32–35 L · -16.8 vs form (-3.9 vs the field)
UCLA · undersold -21 vs Oregonvs Wisconsin 17–7 W · -14.2 vs form (-1.3 vs the field)
USC · undersold -17 vs Penn Statedid not play
Texas A&M · undersold -15 vs Missourivs The Citadel 41–0 W · +31.8 vs form (+44.7 vs the field)
Houston · undersold -10 vs Kansas Statevs Oklahoma State 35–32 W · -20.3 vs form (-7.4 vs the field)
BYU · undersold -7 vs Iowa Statevs Notre Dame 28–24 W · -19.2 vs form (-6.3 vs the field)

Baseline: all 33 page teams that played ran -12.9 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: -22.3 vs form over 4, -9.4 against the baseline (-3.0 in the study) · Undersold: -5.5 vs form over 4, +7.4 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.