Dataset: NBA Block Possession Data

About this Dataset

NBA Block Possession Data

Description

Individual player data from all games from start of 1996-97 season to January 27, 2021

Scraped from Basketball Reference using Python (bs4) and created summary columns using Python
Passed all assertion tests during scraping but not verified rigorously (I take no responsibility for screwing up your analysis hehe)
Users can append GAME_ID to "https://www.basketball-reference.com" to get URL of game webpage
Users can append PLAYER_ID to "https://www.basketball-reference.com/players/" + "[first letter]/" to get URL of player webpage
- ex. https://www.basketball-reference.com/players/k/klebima01.html

Limitations

Significant noise from original source, so separation of possessions isn't completely accurate
- Affects pretty much every column
- Accounted for many different situations, but sometimes the play-by-play was just out of order

Credits

If you use this data, I'd love to hear what projects my fellow basketball nerds are up to or even help collaborate! Shoot me an email at kh19@princeton.edu and please credit me with a link to this kaggle or my website. As usual, any suggestions or tips would be greatly appreciated. Enjoy!

Coding Time: 20 hours
Scraping Time: 6 hours

Tables

Games

@kaggle.kenhuang41_nba_block_possession_data.games

11.13 MB
293810 rows
12 columns


CREATE TABLE games (
  "game_id" VARCHAR,
  "player_id" VARCHAR,
  "player" VARCHAR,
  "possession_id" BIGINT,
  "seconds_elapsed" DOUBLE,
  "points_scored" BIGINT,
  "team" VARCHAR,
  "play" VARCHAR,
  "block_play" VARCHAR,
  "eventual_play" VARCHAR,
  "block_count" BIGINT,
  "shot_count" BIGINT
);