Ranking NFL Teams with Neural Nets

Posted on by

If you had to build a neural net model to help you rank NFL teams, what bits of data would you want to include in your input data?

Pass Attempts vs Completed?
Yards Rushed?
Score Differential?
Time of Possession?

What? I’m interested to hear your comments, if any.

  • G

    i will add:
    1. time to lunch the ball from the hand of the quarterback
    2. total time of the play from when the team gets the ball to the tuchdown

  • Bedo

    Some time ago, I had played with this a bit, and had found that it seemed I needed to work starting field position into the mix. I’ll be interested to see what you find as you go along. Keep at it!

  • http://www.neuralmarkettrends.com Tom

    @G: Thanks, I will keep that in mind
    @Bedo: I’m in the process of gathering data and inspecting it. I’ll probably run a clustering algo through it to see if there any interesting relationships. Thanks for input.

  • http://zerocold.wordpress.com Piero

    you may find a good source even written with matlab from Joshua Khan who has developed a neural network approach to nfl forecasting

    http://homepages.cae.wisc.edu/~ece539/project/f03/kahn.pdf

    Piero