The Packard Rankings were created by Erik Packard who is a Professor at Colorado Mesa in Grand Junction. He received a B.S., M.S., and a Ph.D. from Texas Tech University. While he was at Texas Tech he created the algorithm for the Packard Power Rankings. There is now many modifications that have been made for the extensive algorithm. He gets help from three students in a computer science class, Drake Cullen, Nathan Briner, and Derric Loya.
The rankings come from an algorithm with all information from the season with no bias. It uses final games scores from the week, and it also takes in account for homefield advantage. Along with that the margin of victory is also taken into account.
The score prediction of the games are not the exact for the scores but a approximate of what it will be. If the game is predicted to have 23.4 it is a better of an estimate then just 23. The z-score is also calculated when the predictions are being made. The z-score conveys how many standard deviation above or below the average performance that they displayed in that game. The positive would mean that they move up a ranking and a negative one decreased it. If a z-score is greater than 2 and -2, the game will have the greatest impact. After more and more games the z-score will change and you are able to see which games are going to be better matchups and games that will be less good matchups. After the season is complete you will be able to see which game was the best game.
He gets quite a bit of people that follow him saying, “My understanding is that I have quite the following based on word of mouth and occasional emails. I think my predictions often scare people with how close they are.”
He has also started to do College Football and has continued to do High School Football and Basketball along with College Basketball.
