Machine learning techniques applied to NFL offensive play by play data

Using NFL Scrapper in R by Max Horowitz, able to pull play by play data of NFL seasons and games.

Applying the methods of SVM, random forests, and partial least squares, analysis of teams performance can be garnered.

3-13-17 update Uploaded script file, R.data workspace, as well as initial play by play data for 2015 season. The code needs a lot of cleaning up to look better, but will run as is. It is recommended to download all 3 files NFL-ML.R NFLPlaybyPlay2015.csv NFL_workspace-Working.Rdata

as is to run. The code works off the initial 2015 data of Atlanta offensive plays, and then uses models built on 2015 to compare performance of teams in the 2016 season. The NFL Scrapper by Max Horowitz gave the original csv file found on kaggle, and allows for comparisons to the 2016 season very easily.

Much more to come