Introduction to the Topic: Perception is that games on Thursday are different from non-Thursday games. If this is so…
-
How does this manifest on game day?
-
What is different about Thursday games from non-Thursday games?
We examined the following variables for differences on Thursday and non-Thursday games:
-
Travel Time
-
ELO win probabilities
-
Expected Points (added)
-
Run/Pass Completion
Data: The data for this project is owned and maintained by the NFL. Thus, no data is included in the public repository. We used 2 large data.frames: play-by-play (pbp) and games (games). Calls to these data.frames will be seen in our work.
File Descriptions:
-
data_dictionary.RMD: a dictionary that includes all variable descriptions
-
dataframes.R: a file that includes all necessary data-wrangling and data.frame management.
-
calculating_distance.Rmd: calculates and graphs distance traveled by teams
-
Distance and Accuracy.Rmd: calculate pass completion status and compares to travel distance
-
expected_points.R: examines expected points added (EPA) on Thursday and non-Thursday games.
-
nfl_538_team_codes.R: includes ELO ratings for teams and replicates calculations
-
rankings.R: calculating rankings for teams before games
-
Scores.Rmd: function for calculating final score and margin of victory for all the games of a given team in a given season
-
travel_permutation.Rmd: permutation tests for difference in travel times for different variables
-
travel_vs_margin_of_victory.R: compares travel time and margin of victory for teams
-
vis_team_strength.Rmd: calculating a strength metric for visiting teams utilizing ELO rankings
-
win_travel_permutation.R: permutation test for difference in travel time and win status