The New York Islanders entered this season with a seemingly clear top six defense corps. The seventh man was clearly Brian Strait, with the 8th being one of Griffin Reinhart or Matt Donovan (to the disappointment of a few fans). When all six guys were healthy, it seemed clear who would be playing on defense and who would be scratched. There was no confusion anywhere (ignoring the Strait/Donovan debate).
Since the all star break however, this has changed.
Starting with a Lubomir Visnovsky injury, the Isles have spent all but two of their games with Brian Strait in the lineup instead of Visnovsky or Calvin de Haan - a few of those games with both de Haan and Visnovsky listed as "healthy" scratches. To say that has confused fans both here at LHH and elsewhere is an understatement.
Is there any justification for this? Is this as bad as it sounds? Well, one statistic I've seen used to justify this decision has been +/-, particularly since Jan. 1, 2015. Strait has a higher +/- than Lubomir Visnovsky for the entire season, with Lubo and Hickey in fact being negative despite the Islanders' success -- and some people note that Strait has been very positive since the start of the new year.
There are a few obvious reasons why this is bad reasoning, of course. For example, Brian Strait's positive +/- right now is due to being on the ice for three shorthanded goals, which you'll note he had nothing to do with (and Visnovsky doesn't play PK). Playing PK for a team with a shorthanded threat makes Strait's +/- inflated - he takes no damage to the stat for when he gives up a goal on the PK, but gets a plus for when his teammates get a goal. This wouldn't be a negative if Strait showed any signs of being a good PK guy, except well, Brian Strait allows the most shots on goal on the PK of any Isles' D-man.
The other is that this statistic doesn't hold up when you take a closer look into how it's happened. Below I've included relevant statistics for each Isles defesnemen from Jan. 1 to Feb. 20 (not counting Saturday's shootout loss vs. the Capitals).
Table 1: Relevant On-Ice Statistics at 5 on 5 for the Islanders D since the new year, courtesy of http://war-on-ice.com/.
CF% = Corsi % (also known now as SAT%) - % of shot attempts taken by the Islanders when each player is on the ice.
FF% = Fenwick % (also known as USAT%) - % of unblocked shot attempts taken by the Islanders when each player is on the ice
SCF% = Scoring Chance % = % of scoring chances taken by the Islanders when each player is on the ice
CA60, FA60, SCA60 = Amount of shot attempts, unblocked shot attempts, and scoring chances against respectively allowed by the Islanders per 60 minutes with each player on the ice.
GF% = % of goals scored on the ice by the Isles while each player is on the ice at 5 on 5.
The above stats basically break down what happens at 5 on 5 (so not talking shorthanded or power play situations here) with each D-man on the ice in terms of different types of shots, and then the results of those shots. The key for Strait is that GF% - Goals Fielded % - which is basically just +/- expressed as a percentage and ignoring shorthanded goals.
Looking at that, Strait looks great since the start of this year and Lubo looks terrible (while Hickey also looks poor for obvious reasons). But this is telling us nothing our +/- statistic from above already said - it is the other numbers explain WHY that is.
Simply put with anyone but Brian Strait on the ice, the Islanders since Jan. 1:
A. Out-attempt their opponents
B. Get more unblocked shots taken at the net than opponents
C. Get off more shots from the scoring chance area (in close) than opponents.
These numbers are all really good too - the Isles are dominating in shots and chances under the original six Isles D-Men. But with Strait on the ice, the Islanders are getting significantly outshot and are even getting outchanced! (although not out-chanced by a lot). He's the ONLY defenseman for whom this is true.
In fact, if you look just at shots against -- so the DEFENSE part of these numbers -- the numbers get more ridiculous: Brian Strait is allowing significantly more shot attempts, unblocked shots, and scoring chances on net against than any other Islanders D-Man. And in fact, Visnovsky LEADS THE TEAM since the start of January in preventing scoring chances on net, with Calvin de Haan in SECOND.
Yes, those are the two guys scratched for Strait.
So why is Strait's GF% so high and Visnovsky's so terrible? Well, the answer is further to the right on this table: with Brian Strait on the ice, the Isles goalies are saving 93.3% of 5v5 shots, an elite number, and Isles shooters are scoring on 11.8% of 5 on 5 shots, again an elite number. Are any of these things likely caused by Brian Strait? Not particularly likely - I shouldn't need to convince any of you it's not likely Brian Strait is causing Isles forwards to shoot at a higher clip.
How about save percentage? Well again, remember what we learned above: with Strait on the ice, the Isles see MORE dangerous chances with him on the ice than with him off of it. And as a percentage of shots allowed, Strait is indistinguishable from Visnovsky - Visnovsky allows chances on 50.3% of shot attempts against, compared to Strait's 49.9% - a miniscule difference.
And yet the chances against Lubo have gone in at an insanely high rate compared to Brian Strait. The only explanation is simple: Strait has been bailed out by great goaltending, whereas Lubo has had zero goaltending luck whatsoever - a .809 SV% against is so low it is laughable. This luck is almost certain not to continue, but for this year, it has made Visnovsky's +/- look awful, despite him being amazing at preventing the best offensive chances.
In short, yes the original thoughts of the Isles and their fans were correct - there is a clear top six for this team. And there is a clear drop off from those six to Brian Strait.
There is no way Brian Strait should be playing over a healthy member of the top six, and certainly not getting major major minutes as he has the last few games.