Prediction
Week Two Graded: The Model Went 8-7, and 0-3 Against the Favorite
Husker fans, here’s how the model’s Week Two Big Ten board graded out.
8-7 straight up. 6-9 against the spread. Week One was 12-2.
The scorecard
| Game | Model said | Actually | |
|---|---|---|---|
| Oregon at Oklahoma State | Oregon 38-10 | Oklahoma State 39-31 | ✗ |
| Louisiana at USC | USC 41-14 | USC 49-30 | ✓ |
| Penn State at Temple | Penn State 35-17 | Penn State 27-9 | ✓ |
| Utah State at Washington | Washington 35-16 | Washington 16-14 | ✓ |
| Bowling Green at Nebraska | Nebraska 38-10 | Nebraska 56-7 | ✓ |
| Iowa State at Iowa | Iowa 30-14 | Iowa 16-13 | ✓ |
| Eastern Michigan at Michigan State | Michigan State 31-20 | Michigan State 35-7 | ✓ |
| Maryland at UConn | Maryland 31-21 | Maryland 38-14 | ✓ |
| Duke at Illinois | Illinois 28-23 | Duke 31-27 | ✗ |
| Mississippi State at Minnesota | Minnesota 28-24 | Mississippi State 38-13 | ✗ |
| Ohio State at Texas | Ohio State 28-23 | Texas 24-23 | ✗ |
| Wake Forest at Purdue | Wake Forest 27-24 | Wake Forest 38-36 | ✓ |
| Oklahoma at Michigan | Oklahoma 24-21 | Michigan 17-10 | ✗ |
| San Diego State at UCLA | San Diego State 24-23 | UCLA 28-10 | ✗ |
| Rutgers at Boston College | Rutgers 30-24 | Boston College 28-21 | ✗ |
Average score error: 9.37 points a team. Season to date: 24-11 straight up, 17-17-1 against the spread.
Against the favorite
The betting favorite won 11 of these 15 games.
The model picked against the favorite three times — Rutgers, Ohio State and San Diego State. All three lost.
On the final margin, the line was closer in nine games and the model in five. On Nebraska the two were identical. Average margin error: 14.0 points for the model, 12.3 for the line. In Week One it was 14.5 and 13.6.
The three picks against the favorite
San Diego State at UCLA had the widest gap on the board, 13 points: the model had San Diego State at 51%, the line had UCLA by 12.5. UCLA won by 18.
Ohio State at Texas: the model had Ohio State by five at 63%, the line had Texas by 1.5. Texas trailed 23-3 going into the fourth quarter, scored 21 straight and won 24-23. The pick lost straight up and covered against the spread.
Rutgers at Boston College: the model had Rutgers 30-24, the line had Boston College by 3.5. Boston College won 28-21.
Where the model and the line were both wrong
Four misses came in games where the model and the line had the same winner.
- Oklahoma State 39, Oregon 31. The line had Oregon by 23.5. The model had Oregon at 95%.
- Mississippi State 38, Minnesota 13. The line had Minnesota by 1.5. The model had Minnesota at 60%.
- Michigan 17, Oklahoma 10. The line had Oklahoma by 5.5. The model had Oklahoma at 55%.
- Duke 31, Illinois 27. The line had Illinois by six. The model had Illinois at 64%.
Where the model was closer than the line
- Penn State 27, Temple 9: the model had Penn State by 19, the line by 23.5. The final margin was 18.
- Washington 16, Utah State 14: the model had Washington by 19, the line by 26.5. The final margin was two.
On Nebraska, the model projected 38-10, a margin of 29.5 — the same as the line. Nebraska won by 49. My own pick, graded, is here.
Confidence labels
Every pick carries a label — strong, solid, lean or toss-up. This week:
| Label | Straight up |
|---|---|
| Strong | 5-1 |
| Solid | 2-0 |
| Lean | 0-3 |
| Toss-up | 1-3 |
On the season, strong picks are 17-2. The model’s own season check reports the labels out of order: lean picks are 0-3 while toss-ups are 4-5, a gap the check flags as larger than chance. Both samples are small.
Next
Nebraska hosts North Dakota on Saturday at 6:15 p.m. on BTN. That game isn’t on the model’s Week Three board.
Husker Nation, the Week Three Big Ten board comes before Saturday.