This R project analyzes NFL player performance data using multidimensional scaling (MDS) techniques to identify player similarities and clustering patterns as outlined by Hastie in ESL and covered in the statistical learning course at Imperial. The analysis includes data cleaning, statistical processing, and 2D/3D visualization of player positions based on performance metrics.
Key components:
- Data retrieval using
profootballrefpackage - Distance matrix computation with Canberra distance
- Classical and non-metric MDS implementation
- Interactive 3D visualizations with position-based coloring
For detailed methodology and results, see fantasy_football_player_analysis.pdf.