clock menu more-arrow no yes mobile

Filed under:

Season in Review: Islanders advanced stats like Moulson

Sean Bergenheim draws penalties in bunches.
Sean Bergenheim draws penalties in bunches.

Don't worry, this site isn't turning into a daily stats dump. But I wanted to get some reference points out there when talking about the season that was and the decisions to be made in the summer that will be.

If you don't like "advanced" stats, I understand. But try to look at them as just another angle, another data point, a set of impartial eyes. Hockey is much less conducive to isolated event tracking the way baseball is -- I suspect the traits I like about this sport are the ones that make it so hard to quantify -- but there is still something to be gained from the many attempts to quantify the individual influences on what happens during this chaotic team game.

To that end, here is a compilation of how the Islanders finished according to a few advanced stats/formulas that I like to keep an eye on. I'll give a few words (mostly: links to actual explanations) about each stat category, but take them as you will. (Generally, people use them to suit their own biases, but there you go.) Suffice to say these stats are generally high on Mark Streit, Frans Nielsen and Matt Moulson; they are mixed on Kyle Okposo's year -- which is interesting because most of us see him as one of the best-working, most valuable pieces on the team; but he had a poor plus/minus and some brutal luck this year. And goals, by any measure, are the name of the game.

In virtually every case, you can find one category that praises a player, and one that makes you question that praise. Which is why I like them. (I don't tend to trust people who are certain about everything in this world.)

First, let me say this: Some people treat these as end-all be-all. Some try to make them the basis for predicting future production. I find predicting future performance folly -- whether luck or improved fitness/development, there are always unknown future variables. I treat them as a nice composite to help explain the year that was, and then make educated guesses at what might come in the future.

Just like with plus/minus -- except to a smaller degree -- I'm wary of using any of these stats to compare Islanders with players from other teams. But as comparisons within the team, to get a look at who might be more valuable than we realize, or used in a way that explains other ugly numbers, I think they're excellent.

Brief explanations and links for each category are below the table. The players are ordered by their GVT rank, but that was an arbitrary choice. I boldfaced the best marks and a few noteworthy low marks in each category -- excluding guys who only played 20 games or less.

2009-10 QoC
Pen. +/-
BHT Rating Corsi QoC
Mark Streit -.003 -12 .44 .548 10.7
Matt Moulson .024 7 .68 .518 10.2
Frans Nielsen .031 2 .32 .625 7.9
John Tavares .012 7 -.43 .490 6.9
Blake Comeau -.032 -3 -.20 .237 6.5
Josh Bailey .001 4 .70 -.119 5.7
Rob Schremp -.045 5 .15 -.229 5.3
Trent Hunter .005 E .85 .330 4.7
2009-10 QoC
Pen. +/-
BHT Rating Corsi QoC
Kyle Okposo .041 12 -1.11 .516 3.4
Jack Hillen .017 -12 -.25 .635
Freddy Meyer -.045 -9 .34 -.180 3.3
Sean Bergenheim -.012 10 .43 .644 2.8
Jeff Tambellini -.099 -1 -.03 .112 2.4
Jon Sim -.037 13 35 -.007
Andy Sutton .026 -21 -.10 .726 2.3
Richard Park -.009 E -.09
Dylan Reese .050 -3 1.93
Andrew MacDonald .025 -3 .85 .512 1.4
2009-10 QoC
Pen. +/-
BHT Rating Corsi QoC
Bruno Gervais -.037 -8 -.58 .222 .8
Radek Martinek .054 -2 0 1.595 .7
Dustin Kohn -.110 -1 .47 -.796 .6
Doug Weight -.076 -1 .19 -.093 .5
Dwayne Roloson n/a n/a n/a n/a .4
Jesse Joensuu -.064 1 3.20 -.503
Matt Martin -.007 -1 -1.66 .142 0
Mark Flood .047 0 -4.42 .157 -0.4
2009-10 QoC
Pen. +/-
BHT Rating
Corsi QoC
Rick DiPietro n/a n/a n/a n/a -0.4
Tim Jackman -.032 -3 .03 .779 -0.4
Joel Rechlicz
-.103 -1 -6.09 .849 -0.4
Trevor Gillies
-.060 -7 -1.96 -.912 -0.6
Brendan Witt
-.101 -4 -2.14 -.422 -1.7
Nate Thompson
.069 -3 -1.24 1.156 -2.3
Martin Biron
n/a n/a n/a n/a -5.5

Quickly, a point about why these aren't so helpful in small samples: Notice Joensuu, who played 11 games with about 11 minutes of 5-on-5 per game, has a team-high Behind the Net Rating, as well as a team-worst Corsi relative to QoC. Figures for guys like Flood, Rechlicz, even Reese should be disregarded or treated with massive caveats.

Tom Awad's GVT -- Importantly, this is a cumulative stat that tries to equate to baseball's "value over replacement player" (VORP). So it means actual contributions to actual wins -- including shootout wins. If a player doesn't take shootout shots, he's already had fewer opportunities than his teammates who do. Unlike the +/- and Corsi figures, which try to evaluate players based on 5-on-5 play (of which there is a far larger sample of data to mine), GVT includes all situations.

Here are one, two, three explanations for GVT (Goals Versus Threshold). Or, to quote Awad directly and concisely:

GVT is a measure of observed performance, not implied talent...

So it is a way of describing what has happened -- not necessarily saying this will happen again.

Also: Forwards and defensemen make different contributions; GVT makes an attempt to evaluate their contributions on the same scale, so you can compare apples and oranges.

Behind the Net's +/- -- It's listed as "Rating" at Behind the Net and is described as +- relative to the team. I tend to call this one "relative +/-" although that could be confusing because there are a few adjusted +/- measures out there. It's Gabriel Desjardins' way of trying to control for teammates. (In other words, the Capitals all have massively positive plus/minuses, but that doesn't mean they're all better players than Mark Streit, who finished at even.)

Quality of Competition -- This isn't a grade or reflection of performance, but rather an attempt to measure the quality of opponents a player was on the ice against. "Tough minutes," as some call them. Essentially, the higher a player's number, the greater the likelihood his coach puts him against the opposition's best line -- and the greater the likelihood that all of his stats, from goal production to plus/minus, suffer. A positive number means he generally faced good players, a negative number means weaker competition, and the possibility that Scott Gordon sheltered him.

Corsi QoC -- Corsi relative to Quality of Competition. You probably have an idea about Corsi by now, but if not read this. Like every attempt to measure hockey, Corsi is imperfect. If you play in lead-holding situations, you will face more attempted shots against than shots for, and your Corsi will be negative. Same if you always play against the opposition's best lines (unless you are the rare player who is better than the opposition's best). This is a Corsi rating adjusted for the quality of the opponents you faced.

Penalty +/- -- The easiest to understand, based on a specific trait: Simply, the amount of non-coincidental penalties a player draws minus what he commits. The data isn't always 100% accurate on this, but it's close enough for a season-ending gander. Obviously, defensemen are always going to be in the negative, while forwards (particularly penalty drawers Okopso and Sim) are going to be well into the positive. I bet you wish your rec league tracked this stat for "that guy" who's the hothead on your team ... unless you're that hothead.

*  *  *

P.S. Thanks to those commenters who occasionally cite these are other figures -- that alerts me to the fact that some of you are interested in them. On that note, if there are categories or other angles you prefer to see over these, let me know.

One category I didn't post above is Quality of Teammates. Which is worth considering; for example, Hunter's high rating may be partly due to playing so much with Nielsen.