Wanna blog? Start your own hockey blog with My HockeyBuzz. Register for free today!
 

When Data Disagrees With Decisions

June 11, 2014, 1:56 PM ET [22 Comments]
Travis Yost
Ottawa Senators Blogger • RSSArchiveCONTACT
I half-joked on Twitter yesterday that at some point this summer, I'm going to have to come up with a mini-series of the five biggest changes the Ottawa Senators need to make heading into next season.

It's kind of hard to do now considering the roster turnover that's expected in the next month or so, but one recommendation I figure to be an absolute lock is something like: "Patrick Wiercioch is pretty good at hockey, so perhaps play him a bit more?" Yes, something along those lines.

One of the things that was never really explained well this season was why Patrick Wiercioch was such a regular in the press box. Wiercioch, save maybe a rough October, was one of the team's best defensemen. Granted, it's not as if he had a high bar to clear here. Just about every guy on the back-end struggled or struggled mightily. So, this exclusion from the lineup on many nights was (a) unwarranted; (b) perplexing.

Without saying as much, some of the objections you could tell the coaching staff had with his game were on the defensive side. Another thing I'm skeptical about. In fact -- and I've said this before -- I thought his defensive play was, if anything, underrated this year. If there was anything that wasn't as good as people would have hoped, it was what he was contributing offensively relative to expectations.

Anyways, a fun Patrick Wiercioch fact: in the last two years, he's one of the league's best players on the Corsi% front. When you look at this stuff, generally it's (a) great players on good teams; (b) average or better players on elite teams; (c) guys who are riding some weird systems manipulation who never see Corsi% translate to Goal%, also known as any New Jersey Devils skater.

To illustrate this, I took the top twenty Corsi% defenders over the last two years (min 1000 EVTOI), and added Goal% and Point/60 columns on the right. After all, Corsi% only matters so much because it tells us what goals do not in smaller samples. If Corsi% isn't translating to Goal%, you start getting a bit concerned.

Here's the table, sorted by Point/60:



This really doesn't provide us with a player v. player analysis -- there's a lot more that's necessary, and as you can probably tell, this is not even close to a comprehensive list of 'best NHL defensemen'. (There are many, many, many great defenders not on this table.) Some of these guys hold a Corsi% advantage because of teammate/pairing effects, some of them hold a Corsi% advantage that doesn't translate to a Goal% advantage, and some of these guys hold a Corsi% and Goal% advantage without adding much of anything on the individual scoring front.

What you can say from this table is that just about every guy here belongs in the league. Most play first or second-pairing minutes. And, New Jersey skaters aside, every one of these players is contributing positively to his team's play.

Analysts spend an awful lot of time picking apart specific deficiencies of players, and to a degree, it's entirely understandable (and welcomed!). But, when it comes down to it, does it really matter what you are doing (or what you are not doing) if it's amounting to such a significant territorial advantage?

Sure, you always want to improve an individual player, and finding a specific deficiency could be great in terms of ironing it out and turning the player into a more dominant force. As an aside, a particularly great example would be Ottawa's own Erik Karlsson. As elite a skater as you'll find at this level, but still has issues with pivoting/turning in the defensive zone. Something like that you investigate and try to fix, because it's only going to help him long-term.

The troubling thing about Wiercioch on this table -- against every other player -- is that the coaching staff apparently sees something, that something being enough to justify keeping him out of the lineup.

Why? I have no earthly idea. Maybe he's not strong enough on the puck at times, or maybe his positional play isn't where the team wants it. But, when he's been on the ice the last two years, the team takes a lot more shots and (surprise!) scores a lot more goals than the opposition. And, the point/60 certainly indicates that he's involved directly with a lot of his team's offensive prowess when they're translating Corsi% to Goal%.

The thing about Corsi% (and long-term Goal%, over longer samples anyway) is that it's the ultimate trump card. When the data starts to refute what people are saying, you have to wonder where the analysis is going wrong. Almost always, it's on the subjective side.

Wiercioch has plenty of areas to improve, but it's incomprehensible that he played just 53-games last year. Even if you think he's only accomplishing some of this because of softer zone starts or softer competition, the fact is he's dominating in the capacity he's being deployed.

Which, of course, is a far cry from many of the other options on Ottawa's blue line.

--
Join the Discussion: » 22 Comments » Post New Comment
More from Travis Yost
» Wrapping Things Up
» Enforcer
» Random Thoughts
» Shot Coordinate Fun
» Any Room?