
Northwestern’s points of margin vs a ranked opponent and vs an unranked opponent
, ,
Against what the season’s ratings expected, Northwestern was 7.8 points over vs a ranked opponent and NaN points over vs an unranked opponent over its last 60 games, 13 games against 47. The plain averages are -16.6 and -6.6, a difference of 10.0. All 5 neutral-site games were vs an unranked opponent. All 22 nonconference games were vs an unranked opponent. All 3 games without Casey Harter were vs an unranked opponent.
Northwestern’s NaN vs a ranked opponent and vs an unranked opponent over expected ranks 1st of the 81 teams with 6 games on each side over their last 60 games.
points of margin a game over expected
A NaN difference in points of margin over expected vs a ranked opponent and vs an unranked opponent is rare: shuffle which games were vs a ranked opponent and it comes up less than once in 100.
A −10.0 difference in points of margin vs a ranked opponent and vs an unranked opponent is unusual: shuffle which games were vs a ranked opponent and it comes up 4 times in 100.
The games
| Side | Date | Game | Result | points of margin | Expected |
|---|---|---|---|---|---|
| Feb 17, 2024 | @ | L | -25 | -16.0 | |
| Feb 20, 2024 | @ | L | -25 | -25.1 | |
| Feb 24, 2024 | vs | L | -14 | -16.6 | |
| Feb 27, 2024 | vs | L | -20 | -26.7 | |
| Mar 3, 2024 | @ | W | 11 | -7.3 | |
| Mar 6, 2024 | vs | L | -6 | -13.8 | |
| Nov 10, 2024 | vs | L | -17 | 8.1 | |
| Nov 14, 2024 | vs | W | 2 | -17.1 | |
| Nov 23, 2024 | vs | L | -25 | -7.9 | |
| Nov 26, 2024 | @ | W | 9 | 14.5 | |
| Dec 1, 2024 | vs | W | 13 | 16.2 | |
| Dec 4, 2024 | vs | W | 8 | 4.2 | |
| Dec 8, 2024 | @ | L | -6 | -24.3 | |
| Dec 15, 2024 | vs | W | 10 | 12.1 | |
| Dec 18, 2024 | vs | W | 3 | -6.9 | |
| Dec 20, 2024 | vs | L | -6 | -1.7 | |
| Dec 21, 2024 | vs | W | 2 | 10.9 | |
| Dec 28, 2024 | vs | L | -19 | -9.4 | |
| Jan 5, 2025 | @ | L | -30 | -24.3 | |
| Jan 8, 2025 | vs | L | -4 | -14.0 | |
| Jan 19, 2025 | vs | L | -5 | -12.7 | |
| Jan 23, 2025 | vs | L | -25 | -12.9 | |
| Feb 2, 2025 | vs | L | -14 | -19.0 | |
| Feb 6, 2025 | @ | L | -13 | -19.4 | |
| Feb 9, 2025 | vs | W | 6 | -3.7 | |
| Feb 12, 2025 | @ | L | -15 | 0.6 | |
| Feb 17, 2025 | @ | W | 10 | -3.7 | |
| Feb 20, 2025 | vs | L | -6 | -16.8 | |
| Feb 23, 2025 | @ | L | -5 | -4.0 | |
| Mar 2, 2025 | vs | L | -21 | -11.3 | |
| Nov 5, 2025 | vs | W | 3 | 15.8 | |
| Nov 9, 2025 | vs | W | 13 | 14.9 | |
| Nov 14, 2025 | vs | W | 15 | 14.7 | |
| Nov 17, 2025 | @ | W | 7 | 3.9 | |
| Nov 21, 2025 | vs | W | 7 | 6.5 | |
| Nov 28, 2025 | vs | W | 3 | 1.0 | |
| Nov 29, 2025 | vs | L | -15 | -4.8 | |
| Dec 3, 2025 | vs | L | -12 | -8.4 | |
| Dec 7, 2025 | vs | L | -9 | -23.1 | |
| Dec 14, 2025 | @ | L | -25 | -10.5 | |
| Dec 18, 2025 | vs | L | -1 | 10.6 | |
| Dec 21, 2025 | @ | L | -13 | 3.3 | |
| Dec 29, 2025 | @ | L | -21 | -22.5 | |
| Jan 1, 2026 | @ | L | -33 | -21.2 | |
| Jan 5, 2026 | vs | L | -9 | -22.8 | |
| Jan 8, 2026 | @ | L | -32 | -27.0 | |
| Jan 11, 2026 | vs | W | 19 | 5.1 | |
| Jan 15, 2026 | vs | W | 5 | -3.7 | |
| Jan 18, 2026 | @ | L | -3 | -22.4 | |
| Jan 25, 2026 | vs | L | -34 | -40.1 | |
| Jan 28, 2026 | @ | L | -16 | -23.4 | |
| Feb 1, 2026 | @ | L | -14 | -13.1 | |
| Feb 5, 2026 | vs | L | -18 | -17.9 | |
| Feb 12, 2026 | vs | L | -22 | -30.1 | |
| Feb 15, 2026 | @ | L | -10 | -6.3 | |
| Feb 18, 2026 | @ | L | -36 | -26.8 | |
| Feb 22, 2026 | vs | L | -27 | -15.3 | |
| Feb 25, 2026 | @ | L | -22 | -28.4 | |
| Mar 1, 2026 | vs | L | -5 | -3.7 | |
| Nov 3, 2026 | vs | W | 4 | NaN |
Method
- Expected is an offense and a defense a team, fit over every game of its season between two teams in the table with the game being read left out, plus 2.7 in 2023-24, plus 2.7 in 2024-25, plus 2.7 in 2025-26, plus 0.0 in 2026-27 for the home team (the edge the fit finds once strength is in), nothing at a neutral site.
- Each expected number is then fit to the season’s games: a game bears out about 96% in 2023-24, 97% in 2024-25, 97% in 2025-26, NaN% in 2026-27 of the difference the model predicts.
- The shuffle keeps the season’s shape: the games are cut in two where one date leaves the least variance, each game is read against its stretch, and the sides are shuffled only inside each stretch.