A fair amount of criticism from New York Islanders fans this year has come from two parts of the team especially - the goaltending (which has been well covered by many many articles) and the defense. I hadn't figured the D to be a problem this year, and don't really think it has been as big a problem as many others do, but wanted to take an in depth look at the D in a post for a while.
The thing is though, is that Defensemen are basically never are on the ice on their own. They are always accompanied by a D partner on the other side, and as such, it's sometimes (especially when a pair plays together constantly) near impossible to separate D partners' performance from one another statistically. So rather than look at each member of the Isles' D separately, let's take a look at the Isles D, PAIR by PAIR.
To do this, I'm going to be mainly looking at two numbers: First, the time each D Pair has played together at even strength - I'm only going to be looking at D pairs with at least 30 minutes together with one exception. We're only looking at EV for sample size purposes and because it's the biggest part of the game. Next I'm going to be looking at the corsi percentage of the Isles while each D pair is on the ice. "Corsi" is simply another word for shot differential, except corsi counts missed shots and blocked shots in addition to shots on goal (in other words, Corsi counts shot attempts, not just SOG). Corsi % is often called a possession metric, because it estimates the % of the time a team has possession of the puck in the offensive zone pretty strongly. A Corsi % of 50% means that the team is even in possession, so above 50% is good, above 55% is elite, 45% or below is really bad, etc.
One final note: These numbers are from stats.hockeyanalysis.com, and are thus two games out of date. My apologies for the incomplete data set, it's the best I have.
Okay enough with the explanations, let's get on with it:
Brian Strait D Pairs:
Strait-Donovan (75:12 EV Time On Ice): 44.0% Corsi
Strait-Carkner (32:47): 46.3%
Strait-Ness (27:20): 41.2%
Strait's injury has cost him ice time so I'm including one sub 30 minute pair for him. In fact none of the D pairs he's been on have had significant minutes at all - so I wouldn't go calling any of these #s definitive or predictive by any stretch of the imagination.
But so far, all three of his pairs have been very poor in keeping possession of the puck. To be fair to Strait, he has played some of the fewest minutes with the Tavares line (26.4% of his ice time is with Tavares), but the same can be said of several other D men with better #s without Strait.
In short, Strait's pairs have not worked at all so far. It's not competition or zone-starts either - a change might seem in order.
Aaron Ness D Pairs:
Ness-Carkner (137:17): 52.1%
Ness-Hickey (30:45): 53.7%
Ness-Strait (27:20): 41.2%
Aaron Ness has also played very little with John Tavares compared to other Isles D-Men. In fact, just 26.2% of his ice time (less than Strait!) has been with JT. Even weirder, the third most common forward on the ice with Ness has been Colin McDonald. Despite this, two of Ness' D-Pairs have been very successful, and the D pair of 55-7 (Carkner) has been very solid in a decent # of minutes. Isles fans like to rag on Carkner for the breakdowns, but what's been underestimated is that with he and Ness on the ice, the Isles have spent a good amount of time in the offensive zone and have been pretty productive - the two allow a lot of shot attempts on their own net, but the Isles manage even more at the opponent's net!
In short you wouldn't want 55-7 on the ice in the last few minutes of a game you have the lead, but overall they were a very underrated pairing. Sometimes a D pair can be good enough on offense to make up for defensive lapses, and this pair was.
Matt Carkner D Pairs:
Carkner-Ness (137:17): 52.1%
Carkner-Hickey (75:54): 52.2%
Carkner-Strait (32:47): 46.3%
Carkner-Donovan (31:13): 50.0%
Speaking of poor misunderstood Matt Carkner, know that he's had the least % of his ice time with John Tavares of any Islander D man (24.9%) - his most common forward linemate is actually Casey Cizikas. Despite that, three of his D pairings have been at least even in possession. Again, Carkner gives up a lot of shots against but with him on the ice, the team is surprisingly good at getting shots at the opponent's net. He's a high Risk High Reward player, as you will. When he's bad, he's really bad, but he's subtly good at times. There's a reason he gets more O Zone faceoff starts than D Zone ones (not enough to truly make that much of a difference, but enough)
Worth mentioning of course is that Carkner is -10 in EV Penalty differential - which is a big negative not shown in the above #s. That might be a worthwhile cost if he could keep up a 52% corsi with a D partner for this team.
Matt Donovan D Pairs:
Donovan-Hickey (93:35): 57.2%
Donovan-Strait (75:12): 44.0%
Donovan-Martinek (69:08) 53.0
Donovan-Carkner (31:13): 50.0%
Matt Donovan is NOT another one of doesn't play with Tavares brigade - 31.8% of his ice time was with JT on the ice. That said, his work with Hickey was basically DOMINANT and his work with Radek was very good as well. Basically the only D partner he hasn't worked well with was Brian Strait - for some reason, playing with Strait tended to negate Donovan's offense (his Defensive #s were fine).
There basically was no reason to send Matt Donovan down - again, he might've seemed like another high risk high reward player at times (although with Hickey, they barely allowed any shots against compared to other D men) - but his rewards more than outweighed the risks. Yeah, the shanked shots looked real bad, but the reason he had so many was because he was quite frequently in the O Zone.
Calvin de Haan Pairs:
de Haan-Hamonic:(107:40): 51.4%
de Haan-Hickey (41:29): 49.2%
So CDH has played pretty damn well so far for the Isles, especially as Travis Hamonic's D Partner. That pair has faced very defensive minutes (de Haan's most common linemate is actually Bailey followed by Frans) and has been solidly above even in possession. CDH-Hickey was less successful in much fewer minutes - one really good game, two poor games - but the 44-3 pairing has worked really well, and probably should stick. Unlike the kids mentioned above, the de Haan-Hamonic pairing actually has worked to suppress opponent shots, despite very defensive minutes.
Lubomir Visnovsky D Pairs:
Visnovsky-Hickey (105:01) 52.6%
Included for completeness sakes, this D pair was very very successful last year and was still quite solid this year....come back quickly Lubo.
Thomas Hickey D Pairs:
Hickey-MacDonald (112:04): 36.8%
Hickey-Visnovsky (105:01): 52.6%
Hickey-Donovan (93:35): 57.2%
Hickey-Carkner (75:54): 52.2%
Hickey-Martinek (62:25): 51.5%
Hickey-de Haan (41:29): 49.2%
Hickey-Hamonic (40:27): 53.0%
Hickey-Ness (30:45): 53.7%
The Isles have played 10 D men this year - Thomas Hickey has managed significant minutes with EIGHT of the other 9 (He has avoided Brian Strait). For some reason, since Visnovsky went down, the team basically kept switching Hickey's D Partner around, and until recently, Visnovsky was still his most played-with partner! Somehow, Andrew MacDonald is now his most played with partner.
And man, what a bomb that d pair is. For all the previous complaints about the poor play of Hamonic-MacDonald (more on that later), Hickey-MacDonald is incredibly worse. 36.8% is worse than any single D Man's play in the last full season by a LARGE margin - it means 2/3 of the play is basically in the Isles own zone. The pair may have played some decently tough minutes in terms of d-zone starts, but they are getting WALLOPED. Enough is enough guys.
Outside of that pair, the other 7 Hickey pairs range from just below even (de Haan) to Amazing, with 6 of the 8 pairs all being very good. There's no reason whatsoever for Hickey-MacDonald to exist in any form when you have all of these far superior options available.
Travis Hamonic D Pairs:
Hamonic-de Haan (107:40): 51.4%
Hamonic-Hickey (40:27): 53.0%
For a shutdown pairing, 45% is very poor - hell that's very poor for a normal pairing, but well, even tough minutes couldn't justify those #s. That's what MacDonald-Hamonic were pulling (much better than Hickey-MacDonald, mind you). The team finally decided enough was enough and switched it up, and both of the other Hickey pairings have been very successful. There was some worry that Hamonic might've been the root cause of the problems with the Isles top D pair, but this suggests otherwise. Too bad it took so long for the Isles to realize this, although they may have created a worse pair in the process of fixing our top pairing.
Andrew MacDonald D Pairs:
MacDonald-Hickey (112:04): 36.8%
Amac has the largest amount of time with John Tavares of any healthy Isles D Man (Lubo actually had more) by a good margin. He's had defensive minutes, but not extreme defensive minutes.
And both of his D pairs have been awful - Hamonic-AMac was poor, and as covered earlier AMac-Hickey is GODAWFUL. AMac is one of two Isles D men not to be in a single positive pairing, which is worrisome - one hopes the team won't be offering him a big contract extension.
Regardless, if there is hope for AMac, it's probably on another D pairing than either of these two (and definitely not with Hickey). They tried AMac-Donovan in preseason, I wonder if they'd ever go back to that if they'd call Donovan up. They almost certainly are going to put Lubo back with Hickey if Lubo returns, so Amac is moving somewhere....
And there you have it, an analysis of every D pairing the Isles have put forth this year. Clearly some have worked and some need to burn and die. Let's hope the Isles are paying attention.