Updates / Notes

No neutral site games this week.

Week 9 Statistics

Home/Away Model:

1. Base Version Record: 39-13
2. Base Version Accuracy: 75.00%
   –
3. Slimmed Version Record: 37-15
4. Slimmed Version Accuracy: 71.15%
   –
5. Agree-Disagree Between Versions: 50-2
6. Disagreement Percentage: 3.85%

Neutral Site Model:

N/A

Combined:

Given there are no neutral site games this week, there is no “combined” model record. All predictions records are contained within the Home/Away Model.

Notable Predictions

N/A

Most Unlikely Outcomes

Western Kentucky over Louisiana Tech (14.20% base, 22.06% slimmed)
Rice over UConn (17.78% base, 14.44% slimmed)

Biggest Nailbiters

TCU vs. West Virginia; 23-17 (TCU Win: 85.80% base, 87.83% slimmed)
Memphis vs. #18 South Florida; 34-31 (Memphis Win: 84.12% base, 77.72% slimmed)¹
Nebraska over Northwestern; 28-21 (Nebraska Win: 87.94% base, 88.23% slimmed)

Random Note

I don’t typically include this as a section, but the Oklahoma State @ Texas Tech game was rated as 99.00% base, 99.34% slimmed in favor of Texas Tech, which is the highest I’ve ever seen the win probabilities for either version of the home/away model.

Running Totals

Home/Away Model:

1. Base Version Record: 189-73
2. Base Version Accuracy: 72.14%
   –
3. Slimmed Version Record: 186-77
4. Slimmed Version Accuracy: 70.72%
   –
5. Agree-Disagree Between Versions: 252-10
6. Disagreement Percentage: 3.82%

Neutral Site Model:

1. Base Version Record: 2-1
2. Base Version Accuracy: 66.67%
   –
3. Slimmed Version Record: 3-0
4. Slimmed Version Accuracy: 100.00%
   –
5. Agree-Disagree Between Versions: 2-1
6. Disagreement Percentage: 66.67%

Combined:

1. Base Version Record: 191-74
2. Base Version Accuracy: 72.08%
   –
3. Slimmed Version Record: 189-77
4. Slimmed Version Accuracy: 71.05%
   –
5. Agree-Disagree Between Versions: 254-11
6. Disagreement Percentage: 4.15%

Full Breakdown of Results

Given that the Personal Plan in WordPress doesn’t allow for plugins to be installed, I can’t embed data tables into these pages without running into a ton of formatting issues.

To see the full breakdown of results, please follow the Google Drive link to the Excel document below:

Full Results

1. Memphis being this heavily favored over #18 USF was so surprising to me that I triple-checked the model outputs. While the Biggest Nailbiters section is typically an indictment of the model, as it represents closer-than-expected games (from the model’s perspective), I actually consider this to be a big success. USF was somewhat of a darling in the public eye, so to for the model to go this heavily against the grain is notable. ↩︎