Forensic Hockey
All methods

Action Impact

Experimental

A derived construct without out-of-sample validation - read as informed opinion.

What it means

Credit for every recorded touch by how much it moved the chance of scoring next.

How it's computed

A Markov chain over play-by-play states; each event is credited with the change in expected goals.

What's validated - and what's not

A descriptive model with no out-of-sample validation. It tracks points closely, so read it as context on how value was created - not as a hidden-value detector.

Coverage

2007-08 on; 2007-09 play-by-play recorded shots on goal only, so those seasons are thinner.

Provenance

Routley, “A Markov Game Model for Valuing Player Actions in Ice Hockey” (2015).

Our Markov chain over play-by-play states. Descriptive - no out-of-sample validation; it tracks points closely.

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.