NHL player value model, 2025-26
Hockey analytics for every active NHL skater. The model finds the five closest comparable contracts in the league, weights them by similarity, and returns a market estimate. Deltas vs. actual cap hits surface team surplus and liability across all 32 rosters.
Cap hit vs predicted value
Each dot is a skater. The diagonal is parity. Above the line, the model thinks the player is worth more than the cap hit; below it, the cap hit exceeds the model's estimate.
Largest gaps in either direction
Top five surplus and top five liability contracts by absolute dollar delta. Cluster label provided for context on the player's role.
Surplus
Liability
Surplus by team
Sum of every roster contract's delta against the model. Sorted from largest surplus to largest liability.
How the predictions are made
Comps-based estimator
For each player the model finds five close NHL contracts, weights them by similarity, and returns the weighted average AAV. No black-box regression layered on top.
K-means plus performance score
Seven role clusters (Elite, Top-Line F, Middle-Six F, Bottom-Six F, Top-Four D, Bottom-Pair D, Two-Way / Shutdown). Within each cluster a z-scored production score ranks the player.
UFA contracts get 1.5x
UFA deals are the cleanest free-market signal we have. When they show up in a player's five closest comps they carry one-and-a-half the weight of an RFA or extension.