Mid-Season Grades: Defense

Evaluating defenseman is always a sticky wicket, as exemplified by the divisiveness of the Norris decision every year. Most statistical yardsticks favor offensive stats by nature of the events they're counting, and even those that purport to capture defense have significant flaws, often depicting defense solely in terms of possession, which tends to give credit to the D for what's predominantly the work of their forwards.

Further, it can be quite difficult to extrapolate individual performance from that of the collective. What we know is that as a whole, our team defense deserves an A. We've got the best PK in the league, the fewest goals-against, and are 8th in fewest shots against. We're pretty much right at the historical league average for defense scoring percentage, so even that facet is at least adequate. Each and every one of our regular defensemen has a positive Corsi ON, so nobody's getting overwhelmingly trapped on their own end. Good job guys, A's all around!

Which is not to say that this blueline squad is without fault. As we'll come to see, we've got a couple guys pulling on the rope, and a whole lot of slackers being carried along for the ride by a good system and team-wide buy-in.

---------------------------------------

The Rubric

Per Phunwin's admission in his Forward midterms, GVT and by extension GVS give an offensively slanted view of a player's talents, and as such I won't be employing it here. I could go into a lengthy unpacking of the defensive component of GVT, creatively named DGVT, but given it's basis in shots we're in essence back to looking at possession in spite of Tom Awad's valiant efforts to separately weight these puck-control contributions by forwards and defensemen.

So rather than boiling offensive and defensive contributions, along with salary expectations, down to a single number, I'll be looking at several stats: firstly Goals Against and Goals For per 60, along with the +/- of the two. Let's pause here. You may be asking yourself "+/-? In a stat post? I thought that was useless crap?" Well, self, you're mostly right. Goal differential in and of itself is a pretty circumstantial number, one subject to the vagaries of ice time, luck and goaltending. However, if we look at goals for and against normalized for 60 minutes of play, we can get a better sense of how players with lesser ice time and with heavy duty minutes compare on an even ice surface. Also, if we go one step further and take a gander at GF/60 and GA/60 when a player's ON the ice vs OFF the ice, we can begin to get a sense of their performance relative to the team around them. In the chart below I've provided the GF and GA ON and OFF numbers (all per 60), and then a +/- of their ON and OFF numbers. For those familiar with Corsi REL, I'm applying the same method for the relative figures - subtracting the OFF number from the ON to give a sense of how much the player is contributing to success in that stat category versus their teammates.

I've also included Quality of Competition and Corsi REL for further sense of the pressures of the player's ice time and their ability to maintain possession, along with P/60. I'm mostly going to throw out zone starts, since as the season has progressed the D has gradually come toward the middle of 50% with one notable exception. All numbers will be 5 on 5. It may make sense to look at shorthanded play when scrutinizing the defense, but our PK has been so good that there's very little variation intra-team to compare and contrast. In this realm, everybody aced with flying colors.

I've assigned no weighted value to each category - the grade results remain subjective, but at least fed by measurable performance.

---------------------------------------

MIDTERMS!

Some notes:

For a baseline measurement in goals for and against, Boston's GF/G is 2.82, and GA/G is 2.11. The reason I've not adjusted the GF and GA numbers for comparison to the average is that we're looking at 5on5 figures normalized for 60, and these per-game tallies include PP and PK influence. Still, keep these numbers in your head as a guidepost for average performance.

QoC is extremely stratified, with Claude sheltering the crap out of his third pair. This pretty much gives you the coach's valuation of his D.

I cheated in one place - the GA REL number is actually subtracting the ON from the OFF, so that positive numbers indicate good work. I just reversed the normal method for the sake of message consistency.

And now we begin:

Zdeno Chara: While his usually expected Norris candidacy may be jeopardized by the fact that he's not showing up on the league leader board in points at the moment, he's up there at #15 in ES, performing at the height of his game with 1.25 P/60. The only thing keeping him from said top ranking is a halfway decent powerplay to boost his numbers. There he has a meager two goals and no assists - he's doing his part at least, if only the other guys would bash in his rebounds. While it's just restating the obvious, the team is much better at both producing goals and preventing them when he's on the ice. With him, they have a 1.63/60 differential, ahead of everyone else by a country mile. Without him, they plummet to almost even, with Chara possessing a 1.56 +/- REL. Given that this measure is more drastic than the difference between our goaltenders' performances, Chara is arguably more critical to the overall defensive play than who's between the pipes. Were this graded on a curve, everyone else would fail. A++++, would buy again

Johnny Boychuk: Ah, the curious case of Mr. Boychuk. Pretty much tied at the hip with Chara, he's heavily influenced by his partner, inevitably facing tough and sizable minutes and getting roughly the same zone deployment and high quality of competition. By and large he does his job playing second banana and mostly avoids getting torched by forwards steering clear of Big Z. Mostly. He has the D squad's best Corsi, so he's driving play pretty damn solidly but at the same time also possesses an iffy GA ON, evidently as a result of his ice time away from Chara. He's slightly underperforming the team average in both offense and defense. There's not much difference in offensive production whether he's on or off, as result of the fact that he's producing VERY little himself - even in comparison to his past showings - though the goal prevention markedly improves when he's on the bench, per his GA REL. In fact, Chara sees a 20% positive swing in +/- ON differential when away from Boychuk. Basically, I can't give him too good a grade for looking over his lab partner's shoulder. D+

Dennis Seidenberg: Here's that exception to the rule on zone starts. While he's beginning to move to the middle in his deployment, up until last week he was seeing more than 60% of his starts in the defensive end. With still just 44% Ozone starts he's the Chris Kelly of the blueline, rolled out in defense-first trouble spots. As a result, he has the worst Corsi on the team, though his Corsi ON is commendably still in the black. As with Corsi, the goal output improves modestly when he's off, mostly because he's in the wrong end to start with when he's on. Yet even so, the team stays on the right side of the balance sheet with Der Hammer on the ice, holding the second best on-ice goal differential on team behind you know who, and the second best GA REL to only McQuaid, who has much lesser competition. It's actually quite impressive that he comes away with a +/- REL of just below even given the difficulty of his deployment. B

Dougie Hamilton: By offensive measures, Dougie is a natural. In his rookie season playing second-unit minutes, he's just a single point behind Chara. GREAT SUCCESS you say? Not really from a defensive standpoint. While his p/60 output is good, the team as a whole actually plays better with him off the ice. His GA ON and GF ON cancel each other out, so he's a net zilch at ES, good for lowest on the team. (Again, nobody's in the red though, so YAY TEAM!) When he's off the ice, our differential is nearly a full goal better per 60 minutes. Claude's giving Dougie every chance to succeed, handing him easy competition and big minutes, and giving Seidenberg all of his defensive duties. He's the very picture of an offensive defenseman undergoing the defensive growing pains of transition into the big-league, and his team worst +/- REL shows it. C-

Andrew Ference: The one thing Ference has going for him is that he's been one of the better point producers, per-60 outproducing all but Chara and Hamilton in his limited minutes. However, that modest compliment is where the praise ends for the aptly placed number 6 defenseman on our list. Behind Hamilton he has the worst +/- REL differential. There's virtually no change in the team-average scoring rate with him around, but when he's on the opponents are scoring more. Team-worst more. Relative to Quality of Competition, he has the second worst Corsi of the bunch. Just for keeping his head above water in his spectacularly underwhelming contract year, I'll spare him the failing grade, but here's where the upgrade needs to happen at the deadline. D-

---------------------------------------

In summary, Chara carries this team defensively. Duh. He's the only player that makes a markedly positive offensive and defensive impact. Seidenberg is a valuable force in our own end. McQuaid is just good enough to handle his third pair job well. Ference is trouble waiting to happen, Boychuk's a puppet controlled by Chara's hand, and Hamilton has yet to find his way around his own end.

Having seen Aaron Johnson about as much as Tim Thomas this season, and with Torey Krug always missing the bus from Providence or something, the remaining defensemen didn't show up for the exam.