DARYL SCORE: Identifying Contractually Efficient NBA Players

DARYL Score is a novel metric that uses salary statistics and z-scores to determine how well a player performs relative to their contract. This metric can help NBA front offices allocate salary.

Comment

Inspiration

I love the NBA, and specifically roster construction. A team's playoff success is often determined by how they optimize their payroll. I was interested in seeing which players were on the most economically-wise contracts for their teams.

What it does

The algorithm calculates z-scores for a player's salary and their on-court performance. If a player has high on-court performance but a low salary, DARYL score would rate the player as high.

How we built it

I used open-source data from Basketball Reference, and FiveThirtyEight. I used Pandas, Numpy and Sci-Py to perform data pre-processing and statistical calculations. For the front-end I was planning to use React and React-Bootstrap.

Challenges we ran into

I ran into the issue of data not being int the correct format or not being fully clean. To use z-scores the data should follow a normal distribution so data transformation was needed.

Accomplishments that we're proud of

Completing the full analysis script and data scrapping all the player images.

What we learned

Some React-Bootstrap and React, a lot about Pandas and Numpy, and I learned a bit about Beautiful Soup and Requests.

What's next for DARYL SCORE: Identifying Contractually Efficient NBA Players

Completing the front-end of the website.

Share this project:

Updates