Data analysis R and Python scripts the results of which have (mostly) been posted to my Twitter (@AG_8). Most of the analysis is done using the nflscrapR and nflfastR datasets, which are two of the greatest free sources of sports data in the history of the world (totally not exaggerating).

nfl_fast_scraping.R - mostly self explanitory, it uses nflfastR to quickly scrape play by play and roster data for NFL games and teams from 2000-2019.

nfl_drives.R - analysis of NFL drive data from 2006-2019. Calculates a new Drive Success Rate metric- counting a drive that increases an offense's win percentage as a successful drive. Found that it is similarly predictive as EPA per play (more predictive than Football Outsider's Drive Success Rate), although it is less stable compared to FO's Drive Success Rate.