[Hoot Gibson]

Here is a screenshot of the very bare bones application. It runs locally on my desktop. This is just the output of the program. As you can see, the rankings still don't pass the eyeball test, even after seeding it with last season's results. After two or three more games, it should look much more reasonable, especially after I add the Ohio and Indiana game scores to the program. Excluding out of state games when ranking teams really penalizes teams like Trinity. The predicted accuracy number should be at least 85%. My goal is 90%, which is on the upper end of Massey's accuracy. The current estimated accuracy is only 71%, which is probably pretty typical for early season computer rankings.

There is no way that I could have written this application and imported the schedules and scores for two seasons without the help of an AI. Also, this application is written in a language called Python, which I have only used a little at least 5 years ago. So, I basically created this application by writing the project requirements for Grok and then testing it and letting it know what was not working. I did figure out how to import the data from the Riherds team pages when Grok was struggling to get it right, but the AI did nearly all of the coding.