• The Devils Greatest Trick Was Turning 70% Of Their Players Into Possession Machines

    by Tyler Dellow • October 1, 2013 • Hockey • 7 Comments

    One of the many unfortunate things about having NHL teams in markets that aren’t rabid hockey markets is that interesting things can happen that slip under the radar. If the Florida Panthers dressed a horse for a game, odds are that we’d never hear about it in Canada. The only reason I can say with any certainty that the Phoenix Coyotes exist is that I watched the city council meeting that gave them all the money they had. And maybe the New Jersey Devils had all sorts of really weird things going on with their Corsi% last year that nobody’s pointed out.

    Actually, that last one’s true. Ten forwards were on the ice for at least 200 open play Corsi% events for the 2011-12 and 2013 New Jersey Devils. I’m going to split them into two groups. You should be able to see why

    Screen Shot 2013-10-01 at 10.08.22 AM

    Screen Shot 2013-10-01 at 10.08.32 AM

    These seven guys all saw massive improvements in their open play Corsi% – huge. Seven or more percentage points each. It’s phenomenal. Zubrus, Kovalchuk and Josefson don’t show anywhere near that improvement. Josefson barely makes my cut-off and Zubrus missed half the year so Kovalchuk was sort of a lone holdout, resistant to this Corsi% tide.

    Screen Shot 2013-10-01 at 10.08.46 AM

    Screen Shot 2013-10-01 at 10.09.00 AM

    The Corsi% didn’t increase because of an increase in the ability to generate shot attempts. Broadly speaking, both groups were pretty much the same, year over year.

    Screen Shot 2013-10-01 at 10.09.19 AM

    Screen Shot 2013-10-01 at 10.09.29 AM

    The SAA/100, a measure of the ability to prevent multiple shot attempts on the same shift, is a bit of a different story. Kovalchuk and Zubrus showed some marginal improvement but my first group of players ranged from marginal improvements to large improvements in this facet of the game.

    Screen Shot 2013-10-01 at 10.09.49 AM

    Screen Shot 2013-10-01 at 10.10.07 AM

    And again, you see with the ratio of shifts with 1+ SAF/shifts with 1+ SAA, there was a massive improvement with my first group of seven and then basically nothing from Kovalchuk, Zubrus and Josefson.

    What we see, with the forwards save Kovalchuk, Josefson and Zubrus, is a huge improvement in open play Corsi%, SAA/100 and SAF 1+/SAA 1+. It adds up to a huge swing in the Devils’ possession numbers. Let’s look at the defencemen:

    Screen Shot 2013-10-01 at 10.10.21 AM

    Screen Shot 2013-10-01 at 10.10.32 AM

    Four guys with big jumps, three guys who were stagnant. In reality, it was only two – Zidlicky’s numbers were far, far worse with Salvador than without him. I think you can isolate things down to Larsson and Salvador.

    Screen Shot 2013-10-01 at 10.10.47 AM

    Screen Shot 2013-10-01 at 10.10.59 AM

    SAF/100 changes: Basically none.

    Screen Shot 2013-10-01 at 10.11.13 AM

    Screen Shot 2013-10-01 at 10.11.24 AM

    SAA/100 – some big changes for Greene, Tallinder (who didn’t play a ton), Fayne and Larsson; the others were basically flat.

    Screen Shot 2013-10-01 at 10.11.37 AM

    Screen Shot 2013-10-01 at 10.11.52 AM

    And again, as with the forwards, some massive eye-popping numbers for my first group of four in terms of their ratio of shifts with 1+ SAF/shifts with 1+ SAA. Not much in the second group though.

    I can tell you as well that the Devils posted some mindblowing numbers following 5v5 OZ faceoff wins this year – their Corsi% in that situation was north of 85%, the best I’ve ever seen. So what’s going on? I’m not entirely sure. The downside that the Devils had is that they couldn’t score goals or make saves so all the Corsi% in the world didn’t get them over the line for the playoffs. Cory Schneider will probably fix the latter problem; we’ll see if the former was just bum luck or whether there’s something else to it.

    The real question – and one for which I have no answer – is why. Why, when Kovalchuk, Zubrus, Salvador and Larsson weren’t on the ice were the Devils such a dominant team in terms of their Corsi%? How did they suddenly get so good at limiting second and third shot attempts for the opposition? How did they become so fantastic in terms of absolutely crushing the opposition in probability of getting a shot attempt on a shift versus allowing one?

    What you see in these graphs isn’t a usual team – it’s very unusual. Basically 2/3 of the team became elite outshooters last year, somewhat out of the blue. New Jersey’s a quiet team, in that nobody cares about them and they don’t really talk to the media about what goes on internally but it strikes me that there’s a pretty massive story here and that the Devils are quite possibly into some interesting things or have found something that works awfully well. It’s going to be a fascinating thing to follow this year: if the Devils are still crushing the opposition shot-wise, someone really should look into why.

    One other implication of note: is David Clarkson *really* an amazing possession player or is the NHL’s version of a system goalie at the college level? I’m left a bit suspicious by 2/3 of the Devils team turning into possession all-stars in 2013, to the point that I wonder if, like a goalie who plays for a defensive team in college, players can generate big numbers within the context of whatever it is that the Devils are doing but if they get outside that system, they sort of revert back to being whatever it is they were. Alex Ponikarovsky was putting up decent possession numbers in Winnipeg; they went though the roof when he came to NJ. Matt D’Agostini, who was acquired from St. Louis during the season, saw his open play Corsi% go from 50.8% to 58.3% and his golden ratio from 0.99 to 1.34, albeit in limited time.

    There’s something going on here – I’m sure of it. I’d be eternally grateful if someone else would figure out what it is and write a post proving it but, in the absence of that, it’s something worth keeping an eye on this year as, if the Devils experience some success, it may be something that other teams around the league look to figure out and copy. If only to try and get the David Clarkson they thought they were getting.

    Email Tyler Dellow at mc79hockey@gmail.com

    About Tyler Dellow

    7 Responses to The Devils Greatest Trick Was Turning 70% Of Their Players Into Possession Machines

    1. J
      October 1, 2013 at

      It’s the Devils system.

      Pete Deboer is one of the most underrated coaches in the league. His system is absolutely amazing. I’m not enough of an X’s and O’s kinda guy to explain it but it’s predicated on a deep forecheck and absolutely dominates in puck possession.

      Just need guys to actually score..

    2. Mike S.
      October 1, 2013 at

      DeBoer has shortcomings in some of his personnel decisions at times, but I think he is top notch in terms of getting the most out of his players at 5-on-5. Based on their additions and subtractions this offseason, it’s hard to imagine them being anything less than at least an above average possession team. Add in better goaltending and one would think they have a pretty strong team, despite their lack of a star-studded roster. Unless they are doing something we haven’t seen that is specifically driving their shooting percentages down, they seem like a pretty good bet for the playoffs.

    3. DR
      October 1, 2013 at

      Ahh… good old Poni, he always a great possession player, Sundin really missed him in Van.

      It really does seem like it must be a coaching thing – the Anti-Leafs have the Anti-Carlyle

    4. Triumph
      October 1, 2013 at

      Interesting examination to be sure, but you don’t control for score effects here. NJ spent 818 minutes behind and 550 minutes ahead. That’s rare for a team that controls play to this extent with the score tied. I doubt we’ll see this sort of thing again from NJ, though ditching Kovalchuk will help them territorially.

      Josefson and Zubrus both had shortened seasons and both of them played a lot of time with Kovalchuk – Josefson was yo-yoed around the lineup a bunch but he spent 25% of his time with Kovalchuk (and Zubrus), Zubrus spent about 30% of his time with Kovalchuk. My guesses about Kovalchuk is that basically he doesn’t play in the DeBoer system, whether by his own choice or whether the coaching staff tells him to do that, but he was not pressuring the puck on the forecheck in the way his teammates did and he was almost never the first man to the puck on a dump in.

      As for the Devils’ improvement overall, they ditched Mattias Tedenby and Petr Sykora, got Travis Zajac back (which mitigated Parise’s departure), and got a bunch of games out of Marek Zidlicky instead of Kurtis Foster and Matt Taormina. I bet they were this good in the playoffs, too, or at least for the first 16 games of it. They got run in the Finals, but I assume you weren’t paying attention then.

    5. Eggers
      October 1, 2013 at

      Clearly the Devils were so good at preventing multiple shots against because every first SOG ended up behind Moose and Marty.

      And then with the forwards’ inability to score, you can spend all the time you want in the offensive zone! Shoot, miss, shoot, miss, shoot, miss, shoot, miss, as long as you don’t score, you can rack up all the corsis you want!

    6. Triumph
      October 2, 2013 at

      As for your central question – how did NJ prevent so many multi-shot shifts – could it honestly be partly the wretched goaltending? If you allow the first shot to go in, hard to have a multi-shot shift. I also think zone entries and exits would be interesting to look at – anecdotally it seemed like NJ would carry the play in the opponent’s end and then give up a scoring chance and it’d be in the net. But that scoring chance would be better than anything NJ had been generating in that shift because it came off a rush, and NJ’s chances came off zone time.

      Also I forgot to mention in my last post the addition of Andrei Loktionov, who was a Corsi superstar in a way that someone like Jacob Josefson was not, though we’ll see if that persists.

    7. Pingback: Early Week Grab Bag: “I have opinions about everything!” edition |

    Leave a Reply

    Your email address will not be published. Required fields are marked *