Update (1-1-21)

• Updated file naming scheme to exclude date

Update (12-31-20)

• Included a substitutions scraper script
• Rotations are outputting into Excel tables
• Upon running script, rookies are still not yet registered. SHOTS DATA ARE STILL NOT AUTOMATED.
• Added dates to file names to avoid overwriting files every time.

Update (12-20-20)

• Reformatted percentage rounding in team box scores to integer values instead of a single decimal float
• Data files are not up to date with last minor change, but everything else should be the same

Update (12-18-20)

• Fixed team box scores to include 2-pointers field
• Removed back court shots from team shots data
• Included function to sort shot zones in team shots data

Update (12-16-20)

• Automation script done for everything except rotations data

Scraper outputs

The scraper will work game by game and will only save the most recent game.

1. Spurs' traditional player box score outputs - [date]_players_spurs_box_score.json
2. Opponent traditional player box score outputs - [date]_players_opp_box_scores.json
3. Spurs' traditional team box score outputs - [date]_team_box_scores_spurs.json
4. Opponent traditional team box score outputs - [date]_team_box_scores_opp.json
5. Misc team game statistics - [date]_misc_box_stats.json
6. Spurs' shot data by zone - [date]_shots_spurs.json
7. Opponent shot data by zone - [date]_shots_opp.json
8. League average shot data up the most current point of the season - [date]_shots_league_avg.json
9. Spurs game substitutions - [date]_rotations_spurs.xlsx
10. Opponent game substitutions - [date]_rotations_opp.xlsx

How to use scraper

• Download the entire repository
• Run data_scraper/scraper_script.py to gather, format, and output JSON files

Warnings

• If a previous file exists, this script will overwrite the last file.
• The folder creation may not work on Windows because of the path naming notation.
  • If this is a problem, create a data/ folder outside the scraper scripts folder.
• As of Dec 23, 2020, the shot chats are inaccurate because rookie players are not yet registered.

In progress:

• Rotation chart data
• Rotation chart bar graphs

Data file notes:

• The test game is the final Spurs bubble game against the Jazz: https://www.nba.com/game/uta-vs-sas-0021901274/box-score#box-score
• misc_box_stats.json includes asts, rebs, pts in paint, fast break pts, 2nd chance pts, pts off turnovers, and bench points
• players_spurs_box_score.json and players_opp_box_scores.json are the player box scores
• team_box_scores_x.json are the two team box scores
• shots_spurs.json and shots_opp.json include shots data for each team by zones; includes FG made, FG attempted, and FG percentage
  • FG_DIFF is the difference between the shot area percentage and league average from that zone
  • FG_DIFF is calculated by TEAM SHOT ZONE FG PCT - LEAGUE AVG SHOT ZONE FG PCT
• shots_league_avg.json is the league average shot data for the 2019-20 season

Follow-up notes - VLAD, you can ignore the following

To-do

1. Parse rotations data into bins for visualization

Rotations chart:

1. 30 second bins
2. solid bars for the game

env name - nba-scraper