Datasets Maintained by Donghui Zhang
1. NBA player statistics: nba.txt
Description:
Tab-deliminated NBA player statistics from 1946 to 2004. There are 16,377 records. Besides id, firstname, lastname, and year, there are 17 attributes for every player.
Schema:
- id
- firstname
- lastname
- year
- gp: game played
- minutes: minutes played (not recorded until 1951)
- pts: total points
- oreb: offensive rebounds
- dreb: defensive rebounds
- reb: total rebounds = oreb+dreb
- asts: assists
- stl: steals (not recorded until 1973)
- blk: blocks (not recorded until 1973)
- turnover: turnover (not recorded until 1973)
- pf: personal fouls
- fgm: field goals made
- ftm: free throw made
- tpm: three points made (not recorded until 1979)
- fgmiss: field goals missed
- ftmiss: free throw missed
- tpmiss: three points missed (not recorded until 1979)
Here, (id, year) is the primary key. For attributes turnover, pf, fgmiss, ftmiss and tpmiss, the smaller the better. And for the remaining attributes, the larger the better.
Source: downloaded from databaseBasketball.com in September, 2006. In the zipped file, chose player_regular_season.txt.
Modifications:
- Removed attributes team and leag (and aggregated the rest of the data). This reduces the number of records from 19112 to 16377.
- Renamed ilkid as id.
- Removed fga (or field goal attempted). Instead, keep fgmiss = fga - fgm.
- Similarly, removed fta and tpa, but added ftmiss and tpmiss.
Other Datasets ...