[CorsiHockeyLeague]

Originally Posted by Itse

I don't buy that, since what work I've seen basically says "numbers should even out over time", which is a bad assumption.

Okay, but there have been literally dozens of analyses like this:

https://www.tsn.ca/defencemen-and-th...ntage-1.567469

Much of the effort on determining whether or not a defender has an impact on save percentage focuses on year-to-year repeatability. One thing I was curious about was whether any available metric could reasonably forecast a player’s impact on save percentage in the subsequent year.

The answer to that question is an emphatic no.

So, what’s the conclusion? For now, there is simply zero evidence that a player can truly impact his team’s save percentage over long periods of time. It flies in the face of what some may instinctively think (myself included, many moons ago), but the counterargument just has no supporting statistical evidence. (The lone caveat here: in the event that additional data becomes readily available, perhaps by player tracking technology, there may be an ability to uncover some real supporting evidence.)

... that all suggest you're probably wrong about that.

That sentence alone should I think prove that both shooting percentage and sv% are affected by coaching. (If one of them is, then both are.)

On the power play, yes, because you have a particular structure and strategy. At even strength, the impact of that strategy is not anywhere near as significant, because the play isn't consistently taking place in one team's defensive zone. Even strength systems can affect those percentages somewhat, I'm not denying that - for example, theoretically, if your neutral zone trap causes far less controlled zone entries for the other team, they may get fewer chances off the rush, which may lead to fewer royal road crossing chances, which are more likely to score (lowering save percentage). But the correlation is far more polluted with statistical noise, because the variance among teams just isn't there on those types of metrics - and we have shot heat maps and high-danger save percentage stats to tell us if something weird is happening. It almost never is.

I get that, but not really. I mean, they finished miles away from the playoffs, and not just because they melted in the end. They projected to be a below 90-point team well before the halfway mark of the season.

They were over a 50% chance to make it in around the end of February. Hell, they were in a playoff spot, with games in hand.

It just seems to be a very, very low risk move. Even if they do become worse at something they do well now, at the absolute worst it gets us a better draft pick.

I'm more concerned about wasted seasons. I wouldn't want a better draft pick, because a better draft pick means "playoffs missed" again. This team shouldn't be missing the playoffs. I don't think it needs wholesale changes or overhauls. I think it needs minor tweaks here and there to address areas where they fell short.

If the people in the organization do their due diligence and decide that the coaching staff has to change or those tweaks aren't going to be effective, fine, but I also wouldn't expect that you can come back without some tinkering with the bottom 6 in an attempt to bump team sh% up and a reliable backup to play ~25 games.