Sign in Subscribe
New York Rangers Analysis

Is The Rangers Third Period Shell Actually Working?

Is the Rangers third period shell working?

There is a lot of tension surrounding the media and the non-media world in the Rangers community right now. I'm not sure if it's because advanced metrics have never been easier to access or if the Rangers are a really good sample for them -- in terms of continuing to trust players they shouldn't or them being good despite some of the warning signs flashing.

I think one of the most prevalent issues with this standoff is that both sides have started lobbing grenades at one another. That doesn't help anyone. I really am patient with those who come to me with an intent to try and understand the stats -- or at least question them intelligently. When people tell me I'm a moron I get snarky and that doesn't help either. We're all at fault.

It's not the media's job to report on the team the way you want them to. If they want to use the eye test and the eye test alone that's there prerogative. You don't have to read them. If advanced metrics supporters want to throw everything to the charts and ignore most of what happens on the ice that's their choice as well. The Internet is a free place, you don't need to read anything you don't want to. I write the way I think -- I think the metrics need to be blended with the game itself. The stats color in the picture the game sketches for you.

In an attempt to work through some of these issues, I've tried to react to some of the prevalent narratives being outsourced by the eye-test crew. It's not meant to be seen as an attack, but rather as a thoughtful, logical response to show that maybe, just maybe, the stats show something different than people are thinking. Most recently it was my response that Keith Yandle (and to a lesser extent) Dan Boyle have been major problems for the Rangers this year.

Today it's going to be about the excuse that the Rangers' analytics don't matter -- or that analytics as a whole don't matter -- because the Rangers keep winning despite going into a shell in the third period with the lead.

By now you've probably seen it dubbed as "the stat" among the beat reporters. If you haven't seen it, it's below:

The stat, in a lot of ways, is being used to discredit the Rangers advanced metrics -- and this is true of everyone who thinks metrics have no place in the game. It's being used to say a bad corsi doesn't matter -- since the Rangers have a terrible corsi in the third period and still win -- and that it proves Dan Girardi (or anyone else playing big crunch time minutes late) only has a bad corsi because of this strategy or that their bad possession numbers don't impact the team negatively.

You want to know why the Rangers are 158-1-9 when they enter the third period with a lead? Elite goaltending. That's really it. The Rangers shell in the third period actually hurts them, and if it weren't for elite goaltending that stat wouldn't look anything like it does right now. And, to that point, if even 15-20% of these games went in another direction I don't think the Rangers' defense (or the team as a whole) would be looked at so favorably.

There's a few pieces to this, so let's break them down one by one.

That stat comes since Feb. 6th, 2010, right? So here's the Rangers' collective save percentages and shot differential in third periods from Feb 6th, 2010 to this past Saturday's win over Arizona in comparison to the rest of the league:

The above is all situations (even strength, power play and penalty kill included). The below is just even strength:

These charts should scare you. To summarize:

At all strengths the Rangers are below the league average in shot differential in four of the six years (to the tune of a -34.88 average). Henrik Lundqvist and company, meanwhile, have supported those massive drops in third period corsi with herculean goaltending efforts, to the tune of a 2.1% increase over the league average overall and supporting an elevated save percentage all six years. That is not good defense, that is exceptional goaltending.

But power plays and penalty kills might skew those numbers, right? Well, yes, but the results don't get any better for the Rangers at even strength; they actually get worse.

At even strength the Rangers are -43.90 in terms of shot differential against the league and once again got an extra 2.1 SV% compared to the rest of the league. In all but one year the Rangers were below the league average in shots differential and in all six years they were above the league average in save percentage.

This current year -- where most of the angst is spawning from -- the Rangers are getting out-shot -68.27 to the league average and are seeing a 6.1% bump in save percentage against the rest of the league. Those numbers are very simply not sustainable, even for Lundqvist and the sudden god-mode that is Antti Raanta.

Those stats are all score situations, though. Teams, naturally, shell themselves when they have a lead. So let's take a look at what the Rangers are doing against the NHL when they're ahead in the third period. First, all situations:

Now let's take a look at what the numbers are at even strength.

This tells a very similar story for the all score numbers, but the reality is the Rangers shell a lot more than those team do, and again, are bailed out by exceptional goaltending. The Rangers have an even-strength save percentage 2.96 higher than the NHL average over the past six years when leading after two periods. That's the Henrik Lundqvist factor, not exactly the good defense factor.

One final set of charts. Here's the Rangers' respective numbers (just even strength this time) in all those years up against the Stanley Cup Champion team in the regular season.

And here's the Rangers' numbers against the eventual Stanley Cup Champion's numbers in the playoffs only.

Note: The green numbers in save percentage here actually aren't a good thing. It just denotes the team wasn't above the Stanley Cup team's numbers. So in reality, the Rangers being a -16.6 in save percentage isn't good, but because it's below the "average" it gets coded as green.

The numbers get a little better here in terms of possession, but it does prove my point which I will make below. The Rangers lose that edge in goaltending sometimes and can't recover -- you only get so many third periods with a lead in the postseason. This isn't a knock on Lundqvist, either. This happens everywhere. If the Rangers didn't sit back as much -- they did a good job of this the year they went to The Cup -- it would be to their benefit.

So to the argument that will formulate from all this: But it has been sustainable for six years.

There's nothing inaccurate about that argument, either. It has worked for six years, and through 14 games this year it has continued to work. But there is one major difference. There's a very important word stats people use a lot and I think it plays a major role here: Sustainability.

There's no need to beat around the bush: Good isn't good enough for the New York Rangers anymore. The only goal is to with the Stanley Cup. There is no Plan B, there is no other option. Anything else is failure.

When a team has had the kind of success the Rangers have had there's only one way to make sure this golden age isn't just an empty memory, and that's a Stanley Cup parade in New York City. To get there the Rangers need to make sure they're able to sustain their strategies to win games. Yes, this shell has worked in the regular season and it's helped the Rangers amass a spectacular regular season record. But they have no Stanley Cups to show for it.

It's not out of the realm of reality to see this not working at some point in the near future. Or it might work for another few years -- having the best goaltender in the world has its perks, you know. But it doesn't take a lot to tip these scales in the favor of things going badly, and that's a problem for a team who has expectations that are so high. Teams who eventually win the Stanley Cup seem to do a better job protecting the lead by not sitting back as much as the Rangers have been in the past. Maybe there's something to that.

So has this worked? Yes. Is it sustainable? Yeah, but I don't know for how long.

That's the problem.

Read next