Coach rating

Validated model

Tested on data the model never fit, with the accuracy published.

What it means

A 0-100 rating built only from coach signals that repeat year over year and travel with the coach.

How it's computed

Reliability-weighted percentiles: puck control 0.51, discipline 0.38, special teams 0.11.

What's validated - and what's not

Every scored axis passed a split-half repeatability gate; the weights are the measured reliabilities. Signals that failed the gate - including arrival impact - appear as context only and are never scored.

Coverage

Head-coach records 2015-16 on; 100+ games to be rated.

Provenance

Forensic Hockey original - built from official NHL play-by-play and box scores.

A first-party construct. We never adopt an idea on authority alone; the methodology stance is recorded in our research ledger.

Our research ledger - every source we tested, adopted, or rejected →

Tiers are a promise, not a style. Reconciled is checked against the official NHL record; Validated is tested on data the model never fit; Experimental is a derived construct without out-of-sample validation - read it as informed opinion. When in doubt we take the lower tier.