Hello fans and welcome to a supplemental post. Before I get into that, I want to thank everyone who has been sharing this blog with others. The hits are increasing steadily as well as the subscriptions. I cannot say how much I appreciate it and hope that we can keep the momentum going. The posts now automatically post on both Twitter and Facebook so that should help as well.
Normally on Sunday, I post along with the ranking (aka G’s Evaluation) the GCR’s record vs its prediction. I realized during last night’s post (aka G’s Exploration), but did not want to confuse the analysis. So tonight, I’m posting a special G’s Extra that not only corrects my oversight, but also, as an extra, shows the cumulative 3 weeks. I’m thinking about posting all of the GCR predictions each week (Winner/Loser/Pct to Win) but don’t know if you, the reader, think that’s overkill – keep in mind there are over 100 games per week. Comment and let me know if you’d like that add. Truly a reader’s choice. My assumed answer has to be “no”, so if you want to see the whole thing (or a part of it e.g., just the FBS games), comment on the blog.
Now onto last week’s results and yes, they were very disappointing to the parent of the GCR. The clear misses in the 50 decile led to a -3 win result. Thankfully, the 60% decile for the FCS teams kept it from being much worse. I would be remiss not to congratulate Northwestern State (208) winning their first game of the season against Incarnate Word (154) who is now 5-3. The GCR had the Cardinals beating the Demons (how cool is that?) 94.7% of the time, yet they fell 44-41 in OT in front of a thunderous crowd of 3,373.
| Decile | FBS | FCS | Total | Expected Wins |
| 50.0-59.9% | 7-11 (.389) | 7-9 (.438) | 14-20 (.412) | 18.70 |
| 60.0-69.9% | 13-9 (.591) | 19-4 (.826) | 32-13 (.711) | 29.25 |
| 70.0-79.9% | 6-0 (1.000) | 7-4 (.636) | 13-4 (.765) | 12.75 |
| 80.0-89.9% | 5-1 (.833) | 2-1 (.667) | 7-2 (.778) | 7.65 |
| 90.0%+ | 2-0 (1.000) | 4-1 (.800) | 6-1 (.857) | 6.65 |
| Total | 33-21 (.611) | 39-19 (.672) | 72-40 (.643) | 75.00 |
I also realized that I post this table every week (or at least the last 3 weeks), but haven’t done anything to show year (or 3 weeks) to date. Here is the combined table. Next Sunday both of these will be in the normal spot in G’s Evaluation.
| Decile | FBS | FCS | Total | Expected Wins |
| 50.0-59.9% | 29-24 (.547) | 33-24 (.578) | 62-48 (.564) | 60.5 |
| 60.0-69.9% | 40-15 (.727) | 34-19 (.642) | 74-34 (.685) | 70.2 |
| 70.0-79.9% | 23-6 (.793) | 21-10 (.677) | 44-16 (.733) | 45.0 |
| 80.0-89.9% | 17-3 (.850) | 9-3 (.750) | 26-6 (.813) | 27.2 |
| 90.0%+ | 6-1 (.857) | 14-1 (.933) | 20-2 (.909) | 20.9 |
| Total | 115-49 (.701) | 111-57 (.661) | 226-106 (.681) | 223.8 |
Even with the dip last week, the GCR is still a little over 2 games above expectation. As a reminder, the GCR calculates the winner that is designed to simulate games, taking into consideration who is home or away (or neutral) and how the teams perform in those situations. The smallest margin of victory rounds down to 50.0% and the max approaches 100% (shown as >99%). The expected wins column above is simply taking the total number of games in each section and multiplying by the midpoint (55%, 65%, etc.). If the actual result is very below or very above, over time, it would indicate an issue with model. While all models are wrong, and some are useful, the proximity of the actual wins and the expected wins across the spectrum indicates a model that is not yet failing (yes, I’ve had WAY too many stat classes) and therefore, currently, useful.
One last thing, I’m about to post another picture. Thanks to John Luth who watched Iowa (15) shutout Northwestern (133) and forwarded a picture. If you attend a game, please take a photo and send it to me at GetCompRank@gmail.com and I’ll post it and give you credit.
Send thoughts/comments/questions and please continue to share,
Thanks, G
