APBRmetrics

I am currently attempting to perform a study on fouling trends in the NBA as my final project for one of my stats classes at KU. I have my theory all worked out but require help in collecting data. Thanks to Dr. Stephen Ilardi, I learned of this community. I have very little experience with SQL so any help would go a long way.
The nature of my study requires certain player, referee, and game data as it exists at the time of a called foul. I wish to center my data on the player committing the foul (Fouler) and the player whom was the victim of the foul (Foulee). I will collect data relative to the individual players and their respective teams in certain instances. Specifically, I wish to collect the following information as it exists at the occurrence of a foul:
• Minutes played (players)
• Fouls committed/drawn (players)
• Points scored (players and teams)
• Number of fouls called by the reporting referee (players and teams)
o I would be satisfied with simply having the jersey number of the reporting referee and convert these to count data later.
• Time left in Game
• Time left in Quarter
• Time left on Shot Clock
For organizational purposes, I would also like to include:
• Game ID: To identify the individual game in which data occurs.
• Team: Respective to Fouler and Foulee
• Player Number: Respective to Fouler and Foulee
I imagine the data-set to look similar to the following example:



This line of example data illustrates the dimensions of a game between the Chicago Bulls and the Oklahoma City Thunder (identified as 24321). A foul is called with 3:32 left in the third quarter and 12 seconds on the shot clock. At this time, Serge Ibaka (#9) has just committed his 3rd foul the 11th OKC foul. In his 25:42 playing, he has 8 of the Thunder’s 68 points. Of the 4 fouls that the reporting referee has called against OKC, this is the first charged to Ibaka. Conversely, Derrick Rose (#1) has just drawn a foul for the 5th time this game. The Bulls have been fouled 11 times at this point. Currently, Rose has 19 of Chicago’s 80 points. Of the 4 fouls the reporting referee has called in favor of the Bulls, Rose has drawn 2 of them. Rose has played 30:37 at the time of this foul.

Once again, my ability to analyze data out-ranges my skills in data collection and management so I am extremely grateful for any help or advice I receive. If you have any suggestions or instructions in accomplishing this task, please comment or e-mail me at bgolenjr@ku.edu.

My initial thought is that looking at data sorted by Time left on Shot Clock would be most interesting within the framework you listed. By player, maybe by age and average for team and position. Good / weak teams by overall defensive efficiency and specifically by FTA allowed rate, etc. Maybe divided into these bins: first 8 seconds (related to fast break), 9-20 seconds into shot clock (regular half-court action) and last 4 seconds (crunch time).

I am a SQL developer. Let me know if there are any SQL related issues that you are needing assistance with.

Nileriver- That's the thing, I do not have the slightest clue on how to put together this data-set. If I were able to put this together, I could be off to the races putting together a study that I'm sure many of you would find interesting but I simply do not have the data management skills to obtain and manipulate all the play-by-play data I need to fill the table in my original post.

You might have some trouble finding play-by-play information. If you don't need this latest season, you can look to see if this has enough for your study: http://www.basketballgeek.com/data/

I have loaded all of that information into my own database as well as created a few tables for normalization purposes. Go ahead and look at the files and see if it has enough information for what you are trying to do. If it does, I could potentially provide you some data dumps of my tables so that you will not need to create any packages to loop through the files. Furthermore, I could help you write some queries to do the analysis that you are needing.