I’m in need of a passion project that’s a continuous work in progress, and NBA Draft History is that answer. I’ve been wanting to address a data science project that answers a real question, want to do this organically, and explore React a bit more.

My initial thoughts are:

1. Accumulate and organize data
2. Hard code data into an HTML page
3. Improve UX by creating a Single Page Applications (SPA) with React.
4. Separate data from the UI, place in a JavaScript file
5. Create an API for the data (maybe in Rails), and wire the two together.

Background

I’m a huge NBA basketball fan, and often like to talk about my beloved Chicago Bulls with friends. This season, while they’re gaining good momentum and have some sort of future going forward, did not start off well with a 3-20 record. This reflected a change in direction from the previous season – mainly surrounding the breakup of Jimmy Butler, Dwyane Wade and Rajon Rondo. Local media forced a Big 3 label onto them, and their potential was only more intriguing in concept. Bickering, injuries, etc. only led to a first round exit in the playoffs, and GarPax chose to hit the reset button.

So, here we are in midseason. The Bulls have carved out a new identity as a team. They’re entertaining with the new players, more often than not anyway, yet not good enough to qualify for the playoffs and too young to make a championship run. When they were putting a winning streak together, more fans were torn about their success. Many of them were saying to “tank” for the purpose of landing a better position in the upcoming NBA draft. It’s a popular belief that non-playoff teams would luck out and select great young talent to help them to playoff contention. They believe: the higher your position, the better your chances of that happening.

I don’t necessarily believe that’s the case, and wanted to point out some historical data to back up my claim. There’s no need to tank. I should add that a team’s front office should do their due diligence, and hope their selection becomes a great player. It’s simply difficult to predict a young player’s success in the league.

Starting from Scratch

I was going to go through a number of NBA drafts, and settled in on a span of roughly 20 years. I’m not entirely sure how to arrange data, but I’ve deduced a few key points.

• Total picks in the draft
• Superstar or Hall of Famer players and their draft position
• Superstars in Top 3, separate from those in AND outside of lottery
• 2nd Tier star/impact players and their draft position, note those in AND outside lottery
• NBA Players with 10 or more seasons
• Players with 7-9 seasons
• Those players with 5-6 seasons
• Players with 3-4 seasons
• Busts with 2 seasons or less

The division of season count, particularly those with 5-6 seasons, is kind of more arbitrary, but wanted a better visual rather than just clump them with another group. Players with 3-4 seasons only last for the rookie contract or slightly longer. Those with less than 2 are pretty much busts.

Count and Evaluation

Rather than play by strict categorization with absolute values, I will be more subjective in assigning players their status; impact players, in particular. I remember most of the names, and will try to channel my best Jerry West to assess their talent.

Programming Things

This is where it gets interesting. I will initially create a static page to serve as more of a template. As the NBA draft numbers also include calculations, I can make the page dynamic with JavaScript. I envision using one of the frameworks, like React, to make this come together.

I can then devise an external API to separate data, while further strengthening the project as an application.

NBA Draft History

GitHub Repo