What happened: Louis Boulet’s framework separates specific shutdown tools from catch-all expected-goals impact. It combines NHL API statistics with microstats from Corey Sznajder’s AllThreeZones Project.
Why it matters: The approach can distinguish chance suppression, positional control and puck-pressure skills that broader models combine. Those differences could sharpen evaluations of top-pair construction and defensive-award candidates.
By the numbers: The model organizes 10 defensive skills into five pairs and draws from two primary data sources: the NHL API and AllThreeZones tracking.
What to watch: The key test is whether the framework remains reliable across larger tracked samples and different player roles.