Skip to content
This repository has been archived by the owner on Apr 4, 2021. It is now read-only.

saiemgilani/cfbscrapR

Repository files navigation

cfbscrapR [archived]

This Repository is Archived – Use cfbfastR

A scraping and aggregating package using the CollegeFootballData API

cfbscrapR is an R package for working with CFB data. It is an R API wrapper around https://collegefootballdata.com/. It provides users the capability to retrieve data from a plethora of endpoints and supplement that data with additional information (Expected Points Added/Win Probability added).

Note: The API ingests data from ESPN as well as other sources. For details on those source, please go the website linked above. Sometimes there are inconsistencies in the underlying data itself. Please report issues here or to https://collegefootballdata.com/.

Installation

You can install cfbscrapR from GitHub with:

# Then can install using the devtools package from either of the following:
devtools::install_github(repo = "saiemgilani/cfbscrapR")
# or the following (these are the exact same packages):
devtools::install_github(repo = "meysubb/cfbscrapR")

Documentation

For more information on the package and function reference, please see the cfbscrapR documentation website.

Expected Points and Win Probability models

If you would like to learn more about the Expected Points and Win Probability models, please refer to the cfbscrapR tutorials or for the code repository where the models are built, click here

Expected Points model calibration plots

(1.31% 1.15% 0.94% Calibration Error)

ep_fg_cv_loso_calibration_results.png

Win Probability model calibration plots

(0.89% 0.787% 0.669% Calibration Error)

wp_cv_loso_calibration_results.png

cfbscrapR 1.0.5

cfbscrapR 1.0.4

cfbscrapR 1.0.3

This was a big update!

  • Updated expected points models and win probability models
  • Add player and yardage columns to cfb_pbp_data() pull thanks to a great deal of help from @NickTice
  • Add spread values to the cfb_pbp_data() pull
  • Add drive detailed result with attempts at creating more accurate drive result labels
  • Added series and first down variables
  • Added argumentation to allow for San Jose State to be entered without accent into cfb_pbp_data() function team argument.