Welcome!

PURPOSE

I created this system in an attempt to create a rating for the division 1 College Football teams which can be used to predict the exact score of each game. Yeah, I know...stop laughing. It was initially started in 2014 and has run every week of the college football season since. It consists of a page that shows a Prediction Report and a page that shows the Results of those Predictions. It also shows a report that has the major statistics of every team along with the statistics of those team's opponents on a weekly basis.

METHOD

One major rule that I wanted to make sure I adhered to was that my opinion and bias should not be a factor in creating the prediction of the final score of any game. I decided to create a rating that consisted of how good a team's offense was and how good their defense was. And that rating would be calculated based on whatever critera I came up with, but would be applied in all cases without sneaking my own bias into it. I was, and am pretty sure, that filtering in or out all of the different statistics might create a rating system that would account for better accuracy in different matchups, but I know I'm not smart enough to know which things affect the final results best. Therefore I decided to start with using just the final scores of each game to give teams their Offensive and Defensive rating and thus come up with a final score prediction. Amazingly, I am surprised at the accuracy that evolved from just using something as simple as the final score.

HISTORY

The predictions are created based on an Offensive and Defensive rating that all teams have. When the ratings were started in 2014, a beginning rating for all teams was created. Those initial ratings were somewhat biased at the time, based on the conference and how the teams usually did from year to year. I had to start somewhere and so I created the first ratings with almost all teams in the major conferences with the identical rating. The teams in the lower conferences obviously had somewhat lower ratings based on what I thought the scores would be in average matchups between them and the "higher" conferences. Even within a conference I adjusted the initial ratings for a few teams up or down based on how they usually did from year to year.

A LITTLE DETAIL

From those beginnings each game played and the score achieved affected the ratings of both teams, and as the scores rolled in, the Offensive and Defensive rating of every team moved up or down based on how they did as the season progressed. It didn't take long for the ratings to stablize to a reasonable spot for all teams. Every new year I took the previous year's final ratings and used them to begin the new year. Obviously, that doesn't reflect the changes that occurred from the previous year to the next due to graduations, player improvement, coaching changes and many other factors. But since I don't have a way to get that information for all that change, nor would I want to put that much effort into trying to get it, and most importantly I did not want my opinion or bias to taint the predictions, I decided that I would just roll over with last year's ratings. And to my surprise, it almost seems that some of my best weekly predictions were in the first 3-5 weeks of the season.

EVALUATION

In any case, I want it to be understood that these predictions could be spot on or way off. I've had a few games that have predicted the exact score for both teams, some games where I was only off a point or two for both teams, and a lot of games where I was off by a touchdown or less for either or both teams. BUT, there are games where my predictions were so far off that I was mystified what went wrong. Many of those games where I was off by such a long way, I realized that the Las Vegas Sports Books were just as clueless or even more than I. At other times I was way off and Las Vegas was much closer than I was.

WARNING

The point that I want to leave you with is that I honestly don't have a clue if my predictions are going to be close or far off and with my lack of confidence, I HAVEN'T BET A DIME on any of my picks and I would suggest that you don't either. If my information or predictions give you a little more confidence then so be it, but I caution you that I've seen the spectrum of how close or far off the predictions are, and they honestly don't warrant taking any risk.

College Football