Exploration of NBA Player Performance & Salary

A three month project culminating in a 10-minute poster presentation/Q & A session. Through the project, I scraped and wrangled data from multiple sources & generated concise, unbiased visuals in Python to effectively showcase player performance and highlight the cost-effectiveness of team salary spending over time.

NOTE:

• A continuation of the project to build an interactive dashboard can be found here.
• UPDATED INTERACTIVE TABLEAU DASHBOARD
  • Write-up of dashboard development process

Directories

• Data: Contains all data files used for analysis and data scraping code. Data collected from FiveThirtyEight & BasketballReference.com
• Data Sources:
  • RAPTOR
    • Contains several metrics such as RAPTOR, WAR, and PREDATOR to evaluate the performance of players on offense & defense in both the regular season and postseason updated annually.
    • WAR: Wins Above Replacement
    • RAPTOR: Robust Algorithm (using) Player Tracking (and) On/off Ratings
    • PREDATOR: PREDictive rApTOR
  • ELO
    • Game-by-game ELO ratings for each team as well as some metadata about each game played
    • Metadata defined here
    • More Info
  • DRAYMOND
    • Alternative metric for defensive prowess (limited range). Methodology defined here
• final_poster: Contains finalized plotting code, plots, & poster.
• draft_poster: Contains plots and draft poster as well as feedback from midway through project.
• draft_visualizations: Contains data wrangling code as well as draft plots & feedback from early into the project.

Other Files

• Final Write-up: Concise summarization of project and my thoughts on the project.