Fun with advanced NBA stats-how to collect data
[This article was first published on Jun Ma - Data Blog, and kindly contributed to R-bloggers]. (You can report issue about the content on this page here)
Want to share your content on R-bloggers? click here if you have a blog, or here if you don't.
Want to share your content on R-bloggers? click here if you have a blog, or here if you don't.
Ever since NBA introduced advanced data to track player on-court movement, there have been a lot of interesting blogs/news articles to show how data could improve the understanding (or destroy your intuition) of the game. Reading them is a lot of fun. Here is one example. The newly introduce metric called KOBE measures how difficult a shot is. A more recent article on http://fivethirtyeight.com/ praises a player to new high using data science (which is arguably more convincing at the first glance). A simple logistic regression is used to predict an outcome (made or not) of a shot based on the shot distance, shot clock remaining and distance to closest defender. Whether the correlation is strong enough to give a reliable prediction is not given. However, the result is somewhat people who have followed NBA would expect. Curry, amongst Durant, Korver and DeAndre Jordan, is the most efficient shooter in NBA.
While the regression is quite easy to formulate, getting and cleaning the data is not so straightforward. None of the articles I found online (including the above) gives a direct link to the data, or even the details how to get them from nba.com. Interested in doing some interesting analysis myself, I decided to make such a data frame. Finding the correct url could be a little tricky, as you really need to dig into the website. Here is one very useful resource about how to scrape data from a website. Here I get a list of player ID and the corresponding names. Data from 2014-2015 is collected from each player in that list and then combined into a single cvs file. The code is published here. And if you don’t want to run the script but just want to play with the data, I included the cvs file as well. The file size is about 30 MB and has more than 200,000 shot attemps.
Enjoy!
While the regression is quite easy to formulate, getting and cleaning the data is not so straightforward. None of the articles I found online (including the above) gives a direct link to the data, or even the details how to get them from nba.com. Interested in doing some interesting analysis myself, I decided to make such a data frame. Finding the correct url could be a little tricky, as you really need to dig into the website. Here is one very useful resource about how to scrape data from a website. Here I get a list of player ID and the corresponding names. Data from 2014-2015 is collected from each player in that list and then combined into a single cvs file. The code is published here. And if you don’t want to run the script but just want to play with the data, I included the cvs file as well. The file size is about 30 MB and has more than 200,000 shot attemps.
Enjoy!
To leave a comment for the author, please follow the link and comment on their blog: Jun Ma - Data Blog.
R-bloggers.com offers daily e-mail updates about R news and tutorials about learning R and many other topics. Click here if you're looking to post or find an R/data-science job.
Want to share your content on R-bloggers? click here if you have a blog, or here if you don't.