The defensive zone faceoff: more common than the offensive zone start by New Jersey. - Vincent Pugliese
The New Jersey Devils are a positive possession team but they are one of the teams in the league who have started the least in the offensive zone. This post takes a closer look at who it affects, how it can happen, and whether it's a big problem.
Zone starts have stood out to me in recent days and they don't look all that good for the New Jersey Devils. Back on Wednesday, I wrote about the Devils defensemen and their advanced stats. It occurred to me that only two defensemen were starting in the opposition's end of the rink more often than their own in five-on-five play: Marek Zidlicky and Anton Volchenkov. What this meant was that those two were being out there in more favorable situations while the rest were stuck with far more difficult scenarios. This makes sense as Zidlicky and Volchenkov have faced weaker competition than the other four regular defenders. However, seeing Adam Larsson, Andy Greene, and Bryce Salvador all start in their own end less than 42% per Behind the Net got me concerned. If the disparity between those three and the third pairing is that large, then what about the rest of the team?
As it turned out, the disparity expands to the forwards as well. In a way, who's on the more favorable side is what you would expect. Before Friday's games, the following forwards had a 5-on-5 offensive zone start percentage at 50% or higher according to Behind the Net: Matt Anderson at 100% in his two games, Bobby Butler (60% in nine games), Stefan Matteau (56% in nine games), Stephen Gionta (52.7%), Steve Bernier (50.8%), Travis Zajac (50%), Ryan Carter (50%), and Cam Janssen (50% in three games). Given that the CBGB line is a fourth line masquerading as a third line, they should be getting easier zone starts. Matteau is an 18-year old rookie, Peter DeBoer should absolutely give him more . Butler is (and Anderson & Janssen was) part of the bottom six, so they're not going to be out there to start shifts in their own end if the coaches can help it.
The only name among that group that stands out is Zajac. He's the only top-six forward who hasn't been starting more shifts in front of Martin Brodeur or Johan Hedberg than near the opposition goalie. This also means that the majority of the Devils' top players are starting in tougher locations. You can see it all at Behind the Net from Mark Fayne's 49.1% all the way down to the extreme of Adam Henrique's 33.9%. You'll even see some non-top six players on the wrong side of OZS% like Jacob Josefson at 41.7% and Krystofer Barch at 44.4%. Given that the Devils have been a positive team in possession this season, this is disconcerting.
It is absolutely worth noting that Behind the Net calculates offensive zone start percentage by dividing offensive zone starts by total offensive and defensive zone starts. Neutral zone starts aren't included since they're, well, neutral. There isn't a big advantage or disadvantage by starting on either side of the red line. That said, Hockey Analysis does have the team stats for offensive, defensive, and neutral zone starts. Before Friday's games, the Devils took 29.2% of all 5-on-5 zone starts in the opposition's end, 33.4% in their own end, and 37.4% in the neutral zone. In comparison to the rest of the league, they're 26th in offensive zone start percentage, 10th in defensive zone start percentage, and 5th in neutral zone start percentage. So even if we consider neutral zone starts, the Devils are still of the teams in the entire NHL with the fewest shifts in the opposition's end of the rink. Furthermore, the Devils are the worst in that category among all teams with a Corsi or Fenwick percentage over 50. How can this be?
Let's take a step back as far how these zone starts are created. An offensive zone start at even strength can only be caused by a number of events. They include: the opposition goalie freezing the puck, the opposition ices the puck, the puck goes out of play in the opposition end without a penalty (e.g. a deflected shot or pass), the referee loses sight of the puck in the opposition end, and the opposition goes offside intentionally. I'm not sure how to count all of the situations (and I'd love to count icings) but I think it's fair to say that the Devils have committed more these events than their opposition given their current disparity in zone start locations. Players who play more minutes would be more likely to be on the ice and therefore take these starts. Should the Devils be able to change (they can't after an icing), DeBoer is more likely going to put out his better forwards and defensemen for starts in New Jersey's end of the rink. As a result, Ilya Kovalchuk and Patrik Elias get more defensive zone starts than, say, Gionta and Matteau.
It's also worth noting that the Devils have not had as many faceoffs relative to other teams so far this season. While it includes special teams, the real time team stats at NHL.com has the Devils at 349 faceoff wins and 404 faceoff losses in 13 games. That puts them at 753 total faceoffs. After factoring in games played, the Devils have averaged 57.92 faceoffs per game - the third lowest rate in the league. In other words, the Devils' games have been more open and free flowing than most other teams this season. This also means there's fewer opportunities for them to get more zone starts regardless of location. For what it's worth, the highest average is owned by Tampa Bay at 67.15. Those extra faceoffs per game that the Devils aren't experiencing could definitely skew their percentages.
Speaking of skewing, the Devils have only played 13 games. It's entirely possible, if not probable, that the Devils have had a couple of games where they ended up pinned back more often than not for faceoffs. With only thirteen games played, two or three nights can have a bigger impact on the total than others. For example, there have been a couple of games where Jacob Josefson gets moved up in the lineup within a game, usually to take Mattias Tedenby or Stefan Matteau's spot. He therefore takes on more difficult situations, which drastically drops his offensive zone start percentage. If that doesn't continue in future games, then his percentage should become more in line with how he's used regularly. If anything, that makes this worth looking at about a month or so from now again. The results may be rather different which makes this finding not as significant than at first glance.
There is a silver lining to all of this. Corsi and Fenwick events are most definitely affected by zone starts. I went over this with the defensemen, calculating their zone-start adjusted Corsi rates. Larsson's, Greene's, and Salvador's on-ice rates were far better than the non-adjusted numbers at Behind the Net. In applying the same concept to the rest of the team, it means the players have been better at driving the play than we may think. For example, Patrik Elias, Adam Henrique, and David Clarkson have three of the highest on-ice Corsi rates for the Devils per Behind the Net. They are at 19.47, 12.28, and 17.13, respectively before Friday's game. That unit has been excellent at moving the puck forward and creating offense in the face of their opposition. If we factor in their low offensive zone start percentages, then their adjusted on-ice Corsi rates before Friday's games are 21.87 for Elias, 22.56 for Clarkson, and 22.11 for Henrique. Those rates would rank among the very best in the entire NHL. An already excellent line that may become the team's "power line" is even better than we think. Given that most of the Devils have offensive zone start percentages below 50%, then we can assume their on-ice Corsi rates are better than what's listed and that means they've been more effective in 5-on-5 play.
As great as that silver lining may be, I wonder if this is indicative of a larger problem. Again, nearly all of the teams i the NHL who have positive Fenwick and Corsi percentages as a team don't have offensive zone start percentages of less than 30%. Such teams don't have all but one of their significant-minute players well below 50% in offensive zone starts. Yes, it does make the Devils' own team possession stats even more impressive, but if it continues, then I'm not sure how their own numbers in terms of possession or zone start finishes (most of the team does finish better than where they start, which is good to see) don't fall to some degree. Throw in the fact that the Devils haven't been good at winning faceoffs - only Zajac has won over 50% of his draws this season - and it could be a weakness. Maybe not a very significant one; but something, nonetheless.
In taking my own thoughts about how it's only been 13 games, I'm not going to fret too much over the disparity for now. What about you? Are you surprised to learn that the Devils usually start more shifts in their end than in their opposition's? Which zone start percentage stuck out to you more, those with very low percentages (e.g. Larsson, Henrique) or those who actually have them above 50%? Will the Devils be able to get more offensive zone starts in future games or will they muddle along? Please leave your answers and other thoughts about the Devils' lack of offensive zone starts at the player or team level in the comments. Thank you for reading.