ML Lab
Explore pitcher similarity using K-Nearest Neighbors and clustering analysis
Search Pitchers
Number of Clusters
7
Optimal K Selection
Pitcher Clusters (PCA)
Loading visualization...
Archetypes
Methodology
- K-Means Clustering: Pitchers are grouped by pitch arsenal similarity using standardized pitch type percentages.
- Silhouette Score: Measures cluster quality (higher is better, range -1 to 1). Optimal k maximizes this score.
- PCA Visualization: 2D projection preserving variance for visual cluster separation.
- Data: Only pitchers with 500+ pitches in the dataset are included (... total).