downloader.py: This scrapes the data all 5.7 million pitches in the last 8 years from the MLB website. It parses HTML to navigate through pages for each month and day to determine which games were played on which days and then parses the final XML that contains the actual game data with all the pitches thrown in the game. It then stores this information in a mongoDB database in the pitches collection.

download_batter_stats.py: This scrapes from the same website as downloader.py to get information on each batter. We ran this on every year from 2008 to 2015. It parses HTML to look find the batter ID's and then uses to find the url containing the xml data on the batter's season statistics. This information gets stored in mongoDB in the players collection.

update_prev_pitch.py: This iterates through all the already-downloaded pitches in our database and looks up information on the previous pitch, then stores that information directly in the current pitch, so that it will be more easily accessible by our feature extracter.

plotter.py: This uses matplotlib to plot various figures showing our results. It plots based on the saved pickle files created by running grouped_batters.py and player_specific.py

player_specific.py: This files looks through the database to obtain information about a specific pitcher. It determines thier average pitch characteristics and counts how often they use particular pitch sequences. This information is saved in a pickle file so it can later be used by plotter.py. Currently, this is set to obtain data about Clayton Kershaw.