Created on 11th February 2024
•
NBA fans and media members often oversimplify the analysis of a player's scoring prowess by only examining their point-per-game and field goal percentage averages. Although these statistics are informative, they do not capture the nuance of a player's shot distribution. Players tend to favor different areas of the court, and as a result, their shooting efficiency from various locations varies. By creating an intuitive visualization tool, a basketball enthusiast with any technical background (or lack thereof) can explore where their favorite players tend to shoot and the percentages at which they do it. With over a million data points available, the potential for comparison between players and entire teams across the past five seasons is nearly limitless. As a result, the user can leave the site with a detailed understanding of player and team shot distributions.
Because of the amount of teams in the NBA and the amount of players in each team, it was challenging to filter the list of all players for the dropdowns. This is especially difficult when coupled with all the seasons, because players are traded often. Our initial solution involved going through our data points and checking to see what players were in what team. While this worked, the slowdown was very noticeable - it would take some time for the dropdown to load. This was a major bump in the UX, and we needed to solve it. After some group brainstorming, and some trial and error, we were able to come up with a novel data structure configured exactly to our needs. It took some planning out to ensure that it would work properly and could scale properly, but the end result was worth it. We were also able to leverage MongoDB with it, and the implementation resulted in a massive 83,000% speedup.
Tracks Applied (2)
Major League Hacking
Technologies used