Need Bulk Play-by-play from Recent Seasons

[djbayko]

Hey folks -

I'm doing a fun study as a side project, and it requires NBA game play-by-play data. Does anyone know where I can get downloadable PBP data in bulk, without having to copy/past individual games?

I found some older PBP data on the web, and I've used it to prove that my hypothesis has merit, but I'd like to make sure it's updated to the times. Changes in rules, pace of play, coaching philosophy, etc. might have an effect.

Regular Season - MISSING
2016-17
2015-16
2014-15
2013-14
2012-13

Playoffs - MISSING
2016-17
2015-16
2014-15
2013-14
2012-13
2011-12
2005-06

Regular Season - HAVE
2011-12
2010-11
2009-10
2008-09
2007-08
2006-07
2005-06

Playoffs - HAVE
2010-11
2009-10
2008-09
2007-08
2006-07

 

[AlNipper49]

https://www.mysportsfeeds.com/api/feed/sample/pull/nba/2016-2017-regular/game_playbyplay.xml?gameid=20161026-OKL-PHI

 

[djbayko]

https://www.mysportsfeeds.com/api/feed/sample/pull/nba/2016-2017-regular/game_playbyplay.xml?gameid=20161026-OKL-PHI

Site is down. Can you explain what I should expect from this? Based on the URL alone, I'm guessing it's an API that will return PBP data if I provide a "game ID", consisting of the date and ho.me/away team abbreviations. I would have to code something for this to retrieve all 41*30=1,230 games of a season, correct? There is no bulk download option? That's really what I'm really looking for since I don't really have the coding skills and am not interested in investing that much time just for this project. The years I got so far were one click links to a dump of a flat file containing every game

 

[AlNipper49]

you can always just install CURL (assuming you use windows) and export it to csv, it's easier on the linux/mac side

Here is a NFL active roster call I was just playing with.

curl -X "GET" "https://www.mysportsfeeds.com/api/feed/pull/nfl/2017-regular/player_injuries.csv" -u username:password > 2016-2017-activeroster.csv

 

[djbayko]

you can always just install CURL (assuming you use windows) and export it to csv, it's easier on the linux/mac side

Here is a NFL active roster call I was just playing with.

curl -X "GET" "https://www.mysportsfeeds.com/api/feed/pull/nfl/2017-regular/player_injuries.csv" -u username:password > 2016-2017-activeroster.csv

It's not getting it into usable format that's the issue for me (well, that is sort of a problem, but if I can figure out the export, I'm halfway there). It's how to make ~6,150 API calls (41 home games X 30 teams X 5 seasons). I imagine I'd need to code something because I'm not going to sit there for two months doing it one-by-one, but I don't even know where to begin with that sort of thing. That's why I was hoping for some sort of bulk download.

 

[AlNipper49]

read up on the api, there are ways around that. It might not be one, but I think you could do it 30 times (1 for each team) or similar

 

[djbayko]

read up on the api, there are ways around that. It might not be one, but I think you could do it 30 times (1 for each team) or similar

I did, and it looks like it's game-by-game, unfortunately.

Thanks for the tip, but if I can't get bulk data, I'm going to just stick with the seasons that I already have. This project is really only worth a few more hours of my time, not days.

 

[bowiac]

There is no bulk source available for recent seasons as far as I know. The BasketballValue.com files are the most recent publicly available bulk data that I've seen.

You can scrape any number of sites pretty easily, but if you have no coding background, I agree that may take a bit. If you're interested, here's a guide about scraping Basketball-reference's draft data in Python, which you can then apply to play-by-play data.

 

[djbayko]

There is no bulk source available for recent seasons as far as I know. The BasketballValue.com files are the most recent publicly available bulk data that I've seen.

You can scrape any number of sites pretty easily, but if you have no coding background, I agree that may take a bit. If you're interested, here's a guide about scraping Basketball-reference's draft data in Python, which you can then apply to play-by-play data.

Yes, BasketballValue.com is where I got my older PBP files. I downloaded them months ago and lost the source.

I've been toying with the idea of learning how to screen scrape. Perhaps I'll do that later if additional needs arise, but for this one, I believe the older files will be sufficient.

Thank you for your reply! It's not what I was hoping for, but definitely answers my question.

 

[djbayko]

In addition to the older years available at BasketballValue.com, I was able to get the following years of PBP data from someone who had used it for a study. If anyone needs it, let me know.

2013-14
2014-15
2015-16

 

[ShishaTd]

@djbayko
I wrote you a Message
I m doing my Bachelor thesis and it would be so great if you could send me the data!

Thank you in advance!

 

[the1andonly3003]

you can always just install CURL (assuming you use windows) and export it to csv, it's easier on the linux/mac side

Here is a NFL active roster call I was just playing with.

curl -X "GET" "https://www.mysportsfeeds.com/api/feed/pull/nfl/2017-regular/player_injuries.csv" -u username:password > 2016-2017-activeroster.csv

@djbayko have you tried importing this data directly into Excel (Menu>>Data>>From Web)?

 

[djbayko]

@djbayko have you tried importing this data directly into Excel (Menu>>Data>>From Web)?

Getting the data into Excel isn't the issue. The issue is getting PBP data for thousands of games when each of those API calls returns only one game.

In any case, this thread is old. I found some additional bulk data and met my requirements a long time ago. The thread was only revived because another user is looking for similar data. I've PM'd him what I have.