Our methods
Every modeled number on this site carries a tier - a promise about how much it's been tested, not a style. Each links to a page with its definition, formula, coverage, and the honest line on what is and isn't validated. We never adopt an idea on authority alone; what we read, tested, adopted, or rejected is in the research ledger.
Reconciled
(1)Checked against official NHL totals across every season we serve.
Validated model
(6)Tested on data the model never fit, with the accuracy published.
The chance a shot becomes a goal, judged on where and how it was taken.
Methods →The chance the home team wins from this exact game state.
Methods →A skater's even-strength impact on chance rates with linemates and competition regressed away.
Methods →Next-season production projected from recent seasons, aged along the curve.
Methods →A 0-100 rating built only from coach signals that repeat year over year and travel with the coach.
Methods →A rolling team-strength rating updated after every game.
Methods →Experimental
(11)A derived construct without out-of-sample validation - read as informed opinion.
One number for how good a game (or season) was - scoring, chances and play-driving blended.
Methods →Game Score above a replacement-level player, converted to wins.
Methods →Credit for every recorded touch by how much it moved the chance of scoring next.
Methods →Goals a goalie saved compared with what an average goalie would concede on the same shots.
Methods →Which team truly won the nets in a playoff series - and whether that decided it.
Methods →The market value of a player's production minus what his contract pays.
Methods →A pending free agent's likely next-deal term and AAV range.
Methods →The gap between results and chance quality, converted to standings points.
Methods →The chance each team wins the series, from team strength and home ice.
Methods →How production typically rises and falls with age.
Methods →Production while tied or trailing, against the player's own overall rate.
Methods →