Matthew Carrington-Fair, Matthew Epstein, Ben Francis
March 2019

To run the program, it is necessary to have a CSV file with pitch-by-pitch information for a number of games that a single pitcher appeared in. In order to obtain this data, we can use the pull_data.py program, which will collect data for every pitch thrown by a specified pitcher during a full season and save the information in a CSV file. Data is pulled from http://www.brooksbaseball.net/.
In order to import pitcher data, you will need to first install the following libraries:

pip3 install pandas html5lib lxml

Then run

python3 pull_data.py "[bbref_game_log_page]" [pitcher_id]

• bbref_game_log_page should be a URL to a Baseball Reference page containing pitching game log information. The link for Masahiro Tanaka's 2018 game logs, for example, is https://www.baseball-reference.com/players/gl.fcgi?id=tanakma01&t=p&year=2018.
• pitcher_id is a six-digit number which can be used to identify a player. Each major league player has an identification number, which can be found by looking up that player's profile on mlb.com. The player's profile page will have the URL format https://www.mlb.com/player/[first_name]-[last_name]-[ID]. Tanaka's profile page, for example, can be found at https://www.mlb.com/player/masahiro-tanaka-547888, meaning his ID number is 547888.

The progam does very little error checking. In fact, other than checking that two command line arguments are present and that the Baseball reference game log page is not from before 2011 (the site from which the statistics are pulled has data beginning in 2011), no error checking is performed at all. If the command line arguments are incorrect in some way, behavior is undefined. If input is correct, the program will pull statistics for every pitch the specified pitcher threw during the year referenced by the Baseball Reference page and save the information in a file called titled [pitcher_id].csv in the working directory.

To download multiple players' stats at once, use the pull_data.sh script. The script takes a filepath as a command line argument. The file to which the path leads should be a textfile and in this file, each line should have the two arguments that the pull_data.py program takes in: a Baseball Reference URL and a player ID. For example, the contents of an acceptable text file might look like this:

"https://www.baseball-reference.com/players/gl.fcgi?id=hunteto02&t=p&year=2011" 488984
"https://www.baseball-reference.com/players/gl.fcgi?id=tanakma01&t=p&year=2018" 547888
"https://www.baseball-reference.com/players/gl.fcgi?id=kershcl01&t=p&year=2015" 477132

To run the script, use:

./pull_data.sh [filename].txt

To run the program, simply execute

python3 main.py [filename].csv