Goalie interference cannot be written down and everyone keeps trying

The rule asks officials to judge intent, contact and effect at once, then asks a video room to agree.

Read the wording once and it sounds reasonable. Read it twice and you notice it is asking three separate questions and pretending they are one.

Was there contact with the goaltender? Was the attacking player pushed into him? Did the contact affect the goaltender's ability to make the save? Each of those is a different kind of judgement. The first is physical and mostly visible. The second requires deciding whose fault a shove was, which is a defenceman's whole job description. The third requires imagining a save that never happened.

That third question is the killer. An official is being asked to model a counterfactual: would this goalie, in this position, have stopped this puck, if that skate had not been in the blue paint. Nobody can answer that. So what actually happens is a proxy: the video room looks at whether the contact was visually significant, and significance is a matter of taste that varies by official, by arena, and by whether the goal decided the game.

You can see the consequence in how coaches use their challenge. It is not really a challenge on the rule. It is a challenge on the temperature of the room, thrown when a goal is important and the replay contains any contact at all, which most goals in traffic do, because traffic is how goals are scored.

I do not have a clean solution and I distrust anybody who says they do. But there is an improvement available that the sport keeps refusing: make the paint the rule. Any part of an attacking player in the crease when the puck crosses the line, no goal, no discussion, unless he was put there by a defender, and even that exception could be narrowed. It would be harsh. It would also be knowable, coachable, and consistent, and it would end the three-minute conversations that currently decide playoff games.

The objection is that it would remove legitimate goals scored by players fighting for position. True. It would. That is the cost, and I think it is smaller than the cost we are paying now, which is that nobody in the building, including the officials, can predict the outcome of a review before it is announced.

A rule that produces surprise every single time it is applied is not a rule. It is a mood.