The NBA Draft

As a fan of the Utah Jazz, I am particularly interested in basketball players coming into the league. Every team gets two picks each year in a draft, and whichever of the new players they pick will be on their team for at least the next few seasons. The thing that makes this interesting is that the order of the draft is based on how many wins the team had in that season.

For Utah, this will be an important draft The team recently traded away several of its stars and seemingly is trying to reset. This can be a good thing for NBA teams because it means they have better chances at getting a good, young player on their team which tend to stay for a long time. All of that depends entirely on how well the team can choose acurately players that will find success. I would like to find something of an answer to the question: does the success a player finds reflect accurately on the pick they were drafted at?

For this project, I found some useful data on the website Basketball-reference,

which has a page for each year of the NBA draft (like this one). They have a robots.txt file which allows for people to scrape that part of the site with a limit of the rate of requests to the site (1 every 3 seconds), so I proceeded to web-scrape the data for several years in the past.

The process was fairly simple for a single site. In python you could simply use the read_html method from pandas and extract the necessary table. Problems arose with how to deal with multiple years. Luckily, the year was included in the url, so it could be solved with just looping over the year and putting strings together.

After some cleaning, it was all ready to be output in a csv and used. The data and my code to obtain it can be found in this github repo

This data has all the necessities for a basic analysis of how successful players are in the league as well as which pick they were in the draft. It leaves me with several methods in the level of success in the player and about 900 data points. I am excited to start looking at this data to see what the data will tell us!