04 April 2016

PDO, Corsi, the Eye-Test and the New York Rangers

Here is an article from BlueshirtBanter about a way to fix the Rangers and give hope to their Playoff chances.

It's a long article with a bunch of New Stat fluff.
PDO - simply add a team's shooting percentage at even strength to their save percentage at even strength. This supposedly shows how LUCKY/UNLUCKY your team is.

This is wrong. It truly removes the idea that a team and a player in particular sets up genuinely higher quality chances in a given moment. These moments are usually not relying on LUCK at all. It is a targeted, skilled play. And in most goal events, that is the case - the players had clear intention and clear execution.

Go look at Buffalo's second goal against the Rangers, Eichel to O'Reilly. It was a sweet pass and perfect finish. No goalie saves that. No luck involved. Eichel forces Antti Raanta to stay on the post. O'Reilly finishes with a quick shot - it actually should have been placed closer to the post...but the speed still makes it basically unsavable.

Is there luck in sports? Of course. But the reason the guys that are on the ice/field/court/whatever is that they rely less on luck and more on skill.

Is there luck in hockey? Again, of course. And especially for the goalies. These guys are basically guessing where the bullet is going without any time to really watch it go by or hit them. Goalies are relying on good positioning and bad shooting more than anything else. But again, the goalies that are in the NHL are relying less on luck than on good positioning - the only real thing a goalie controls.

Corsi - also simple, this is the percentage of even-strength shots your team took out of both teams' even-strength shots.

The idea is that Corsi can give an indication of who controls the puck the most by assuming the team with the most shots had the most puck control.

Again, a very weak assumption. Not to mention, the team that can create a shot that goes in the goal has better puck control than a team that just takes shots.

Don't get me wrong, in most cases and especially when there is a dominant team, shot counts will usually reflect the domination. But it is far from clearing up the story of what happened on the ice.

Ultimately, I think the point of the article is two-fold.
 - the Rangers need to be lucky! (in other words, put more pucks in the net on fewer shots)
 - the Rangers need to not use the guys that have bad Corsi scores (meaning that while player A is on the ice, the team should get more shots for than against.) This is mainly a knock on Dan Girardi and Marc Staal.

But really, it doesn't say much. How do you get luckier? I don't know and I would assume no one knows...luck is just that, uncontrollable. So no action required to achieve this, I guess.

As for Dan Girardi and Marc Staal, again that isn't an answer. What has changed in their games this year? If nothing, then how were these two such able parts of the past few seasons when the Rangers were able to lock teams down in the third period? I think both have the skillsets to be effective. My hope is that come playoffs, they will lock their games down and return to keep it simple in their own end and keep getting in the way of clean shot attempts.

And as a point about this Corsi stat thing being misunderstood fluff - Dan Girardi is the worst Ranger on the team for Corsi...but his +/- (which is goal differential at even-strength) is top 3 for defenseman (top 3 overall). So I guess he is just really lucky? If so, then maybe we need him on the ice more?

I just get annoyed at this type of dumb analysis.

No comments:

Post a Comment