A web-scraper built with R used to calculate Expected Points Generated for MLS teams using data from WhoScored.com and American Soccer Analysis.
I developed EPG to better measure how much a soccer player contributes to their team. Read more here.
-
Install R, and clone this repo.
-
Install RStudio for an easy-to-use dev environment for R scripts.
-
Install Firefox if necessary. I haven't tested this script with any other webdrivers.
Open up a new Terminal window, cd into this repo, and run java -Dwebdriver.gecko.driver=geckodriver -jar selenium-server-standalone-3.4.0.jar.
Open the script in RStudio. Enter and run each of the following commands in the command-line window to install the required dependencies.
install.packages("RSelenium")
install.packages("XML")
install.packages("plyr")
install.packages("data.table")
install.packages("stringi")
install.packages("scales")
Make sure you have set the current working directory to the repo's directory via Session > Set Working Directory... > To Source File Location, and hit the Source button in the top-right of the Source window on the left-hand side of the RStudio window. This will run the entire script and cache the methods so you can use them from the command line.
retrieveAllTeamEPGs(shouldMergeCSVs = FALSE): Creates EPG files for all teams and stores them under /teams. Calls mergeCSVs to collate the CSV files if shouldMergeCSVs is true. Uses retrieveTeamsEPG() under the hood.
retrieveSingleTeamEPG(team): Takes in a team abbreviation from the teams list in the script, creates an EPG data frame for the team, and stores it as a CSV file under teams/team.csv.
retrieveTeamsEPG(teams): Takes in a vector of team abbreviations from the teams list in the script, creates the EPG frames for those teams, and stores them as individual CSV files under /teams.
mergeCSVs(): Merges all CSV files under /teams into one big file for the entire league, which is saved in the working directory as mls-epg.csv.
comparePoints(team): Calculates the sum of the EPGs for all of a specific team's players and prints a comparison between EPG/EPG per game and Points/PPG to the console.
compareLeagueTable(): Consolidates EPG sums for each club and league standings in one data frame, which is saved as mls-epg-comparisons.csv in the working directory.
Example CSV file for Atlanta United via retrieveSingleTeamEPG('ATL'). CSV files for other teams can be found in the /teams folder.
Example CSV output from retrieveAllTeamEPGs(shouldMergeCSVs = TRUE).
Example CSV output from compareLeagueTable().
Example console output from comparePoints('ATL'):
> comparePoints('ATL')
[1] "ATL Expected Points: 18.4296820313489"
[1] "ATL Actual Points: 18"
[1] "ATL Expected Points per Game: 1.53580683594574"
[1] "ATL Actual Points per Game: 1.5"
[1] "% Difference between Expected and Actual Points: 2.35896668534814%"
See LICENSE for more details.