The Super Awesome Prediction Machine predicted the following before the Ohio State game:
"If I use my standard coefficient of 1.5, I get 48-23 Buckeyes. If I double the coefficient, I get 37-24 Buckeyes. Reading between the lines, the math says that Iowa performed poorly on offense against legit defenses, and while weighting the points more helps the defense, there's just not enough firepower on offense. ... The MOST optimistic model I can arrive at is OSU 31, Iowa 24."
The actual final score was 34-24. If I feed Ohio State's performance against back into the system, Super Action Prediction Machine predicts 37-28. The tighter differential coefficient clearly produces more accurate results and so I'm sticking with it for the rest of the season. For comparison, using a looser coefficient, I get a 9-3 season for Iowa; using a tighter coefficient, I get a 7-5 season (which can only be 6-6 now since SAPM had us beating NIU 32-29). Maybe down the road I'll run a Monte Carlo to identify the optimal coefficient but that requires calculus and I don't feel like doing any.
Anyway, so I updated the engine's data set to include our remaining opponents' performances from this past week and re-ran the model for the remainder of the season. Here's what SUPER ACTION PREDICTION MACHINE has for the rest of the year:
Iowa BEATS Northwestern 32-24
Iowa LOSES to Wisconsin 30-25
Iowa BEATS Purdue 32-20
Iowa LOSES to Michigan 32-31
Iowa LOSES To Nebraska 35-26
Final result: 6-6