clock menu more-arrow no yes

Filed under:

2013-2014 Devils Passing Review: How Frequently do the Devils Generate Offense?

New, comment

This is a look at how frequently New Jersey Devils skaters generate offense in each zone and how that compares with the opposition. Read on for the details.

Brad Penner-USA TODAY Sports

Last time, I broke down how New Jersey Devils skaters contribute to Corsi (all shooting attempts) via their shooting and passing. Now, as skaters play more minutes than others, I reduced their contributions to a rate per twenty minutes figure. However, we can do a little better than that. In this article, I’ll look at a variety of stats and how they relate to a player’s ice time. How many minutes at even strength does it take for a player to generate a shot attempt? A shot? How much faster or slower are they than the opposition? Let’s get to it.

Summer Passing Series Links

Primer

Part One: Efficiency and Winning 

Part Two: The Transition Game 

Part Three: Offensive Zone Analysis 

Part Four: Generating Goals 

Part Five: Corsi Contribution

How Much Ice Time Does it Require to…?

That’s the question I started with on this article. A lot of stats use per-twenty or sixty rates, and I use it as well in my Corsi Contribution article. However, during the season I wanted to know exactly how long it was taking a player to generate offense. Who was making more of their ice time? Let’s first look at the Devils skaters before comparing to the opposition.

Frequency of Transition Offense

In my piece on offense generated in the transition phase of the game, I revealed that teams who control the greater number of shot attempts generated in transition win games more often than those more reliant on offense generated from within the offensive zone. Naturally, I wanted to know how often players were generating these chances.

The charts for Devils defensemen can be read thusly: The x-Axis represents how long it took a player to generate a shot attempt from the defensive or neutral zones (D/NZ SAG); the y-axis represents how long it took a player to generate a shot from the defensive or neutral zones (D/NZ SG); and the size of the bubble represents how many shot attempts were generated (volume).

Def_transition_frequency

In looking at the frequency of shot attempts generated from the defensive or neutral zones (D/NZ SAG), we see that the Devils defensemen generally take anywhere from thirty to sixty minutes to do so. Mark Fayne and Anton Volchenkov are even further behind, while Bryce Salvador lags behind as well, just not as far.

Adam Larsson leads the group at generating a transition shot attempt every 32:49 of even strength ice time. Marek Zidlicky, Andy Greene, Eric Gelinas, Jon Merrill, and Peter Harrold follow Larsson as all are pretty bunched up in the forty-two to forty-nine minute range. So, based on the data from this past season the "defensive defensemen" were least effective in generating offense in transition. That should not be surprising, but it does make you wonder why a team would bother to carry one at all. I’m of the opinion that a "defensive defensemen" or "shut down defensemen" isn’t at all necessary or a good thing to have in today’s NHL. If you don’t excel at obtaining and keeping possession to drive play forward, you’re not helping your team win. Simple enough. I don’t see it as a bad thing that Fayne is leaving the team at all.

Zidlicky and Greene generate the most, but both Gelinas and Merrill tied for the third-most shot attempts with eighteen. Larsson generated only thirteen, but this was more a product of not playing enough. Should his frequency remain the same, or even drop slightly, he’d be among the leaders on the blue line in generating offense from his own end and through the neutral zone. Big hopes for Larsson next season.

Harrold was just behind the two rookies in frequency, though did generate actual shots at a faster rate than all but Zidlicky. As noted in the transition efficiency piece, Larsson’s was not good, explaining why he generates shot attempts so quickly, but shots so rarely. Just behind Harrold was Merrill, Greene, and Gelinas.

Zidlicky, as always, remains dominant by these metrics. And Volchekov, as always, will not be missed.

Using the exact time figures of when things happen is another way of looking at efficiency. It gets you closer to being able to pinpoint exactly how often a player will be involved in a goal.

For example: Zidlicky generated four goals in transition this past season, which translated into six hours, fourteen minutes, and eighteen seconds of even strength ice time per transition goal. As goal sample size increases, I’ll be able to more confidently say what the average amount of time between goals in various phases of the game take place. In fact, once I collect more data on all of these stats, clear averages will form based on how quickly or slowly a team generates offense in hockey’s various phases.

Fwd_transition_frequency

Moving to the forwards, I had to use a bar graph as there were several bubbles at one point in the graph, making it unreadable. The blue bar represents how much time it takes for the player to generate a shot attempt, the green bar represents how much time for a shot to be generated, and the purple line charts the volume.

So, when we look at volume, Dainius Zubrus and Steve Bernier lead the forwards with forty-four and forty transition shot attempts generated. What you may not have expected was Bernier to generate both attempts and actual shots faster than Zubrus does. In fact Bernier generated attempts every 22:25, third-best among the forwards, and shots every 59:46, fourth-best among forwards. Tuomo Ruutu actually generated shots in transition the quickest among forwards at 30:34. I’ve said before that I think Ruutu can be an effective forward in the right role on this team. Ryane Clowe was the next best at 42:46.

Patrik Elias and Adam Henrique had nearly identical intervals between attempts and shots. One thing you’ll also see on the chart is Jaromir Jagr’s only stat that makes him lag behind the other forwards. He is not a transition player; Jagr excels at offensive zone shot-generation (see below), as does Travis Zajac. It’s not that I think they’re poor transition players; it’s that they spend so much of their time in the offensive zone. That’s what they want to do. Both generated thirty attempts in transition, but they couldn’t generate shots as quickly as others. Zajac in particular was well behind most forwards. Something to further investigate perhaps.

Andrei Loktionov was one of the quicker forwards at generating transition attempts, but not as great at generating shots with them. Reid Boucher did okay, generating seven attempts in his twenty games. That production pace would have put him right up with Zajac and Jagr over a full season. He generated shots only a few minutes slower than Elias and Henrique.

Ryan Carter was much better than Stephen Gionta, as he generated a shot in transition almost as fast as Gionta could merely generate an attempt. I understand using Gionta in a depth role, but wouldn’t the team be best served by fitting as many possession/skill players into the lineup as possible?

Frequency of Forechecking (Offensive Zone) Offense

Def_oz_frequency

Back to the bubble charts for the defensemen. Here were truly see Zidlicky in a class of his own. He led the group in how quickly he generated both attempts and shots. Now, that’s not the same as being the most efficient, which Zidlicky wasn’t, but his volume was so high that he consistently lapped the field in generating shots.

In the offensive zone, Harrold featured more prominently by way of frequency, but couldn’t muster up more than thirty SAG. Larsson, the most efficiency forward in the offensive zone, only generated eighteen attempts.

Here is another distinction in a player’s shot-generation: Gelinas was more involved in the transition game at generating offense, but in the offensive zone he did so less than all defensemen not named Volchenkov. This is clearly due to his ability and preference to shoot the puck as often as possible. So, it would be fair to say that in transition Gelinas acquitted himself quite well, but once into the offensive zone, he was only thinking about shooting.

Merrill generated shots faster than all except for Zidlicky. Greene was a bit quicker than Merrill on the attempts, a bit slower on the actual shots, but nearly doubled his volume. As we saw in the Corsi Contribution piece, Greene has a tendency to contribute a good deal more to Corsi by his own shot attempts.

Fayne and Volchenkov each needed more than an hour of ice time to generate a shot in the offensive zone. That’s…poor.

Fwd_oz_frequency

And back to the bar graph for the forwards. Considering the Devils were a strong possession team and generated more of their offense from within the offensive zone as compared to in transition, we expect these numbers to look good throughout the position. Let’s see who really was generating shots and who was generating lots of empty attempts.

Jacob Josefson, Mattias Tedenby, and Stephen Gionta all took the longest to generate shots within the offensive zone: at least forty-five minutes of even strength ice time. To say that another way, Zajac, Elias, Jagr, and Loktionov all generated shots faster than Tedenby and Gionta could generate a simple attempt. Josefson actually generated attempts at an okay pace, but the efficiency wasn’t there.

If OZ SAGE (Shot Attempt Generation Efficiency) correlates strongly with winning games based on last season’s data, then you’d want players who generated shots the quickest. The Devils forwards fell into three tiers: players that generated shots around every twelve minutes, every sixteen minutes, and everyone else that was off that pace took anywhere from nineteen to fifty minutes.

Your top players were Jagr, Elias, Zajac, and Loktionov. The next group featured Ryane Clowe, Bernier, Henrique, Ruutu, and Zubrus. The last group featured some who just missed the pace of the second group (Mike Sislo, Michael Ryder, and Reid Boucher), and those that were far off the pace (Brunner, Carter, Gionta, and the Swedes).

Frequency of Corsi/Shot Involvement

Now, some players are naturally more inclined to pass or shoot, so looking at only the frequency of shots generated by passes might make the shooters look worse. So, in the next charts you’ll see, I took the shot attempts generated and individual shot attempts (iCF) and compared that to how often the player generated or took a shot. What you see is how often the player is involved in a Corsi event and how often that player is involved in a shot on goal.

Def_cc_si

When looking at both iCF and SAG figures as one, we see Gelinas and Zidlicky almost identical in how quickly both are involved in attempts and shots. It clearly identifies Salvador as the slowest to be involved in either on the blue line, slower even than Volchenkov. Greene isn’t far off the Corsi pace of Zidlicky and Gelinas, but both he and Fayne a few minutes behind the shot involvement pace. Merrill and Larsson look a bit slower than most defensemen.

Fwd_cc_si

Jagr again leads the group, though Sislo was a close second in shot involvement pace. Granted (and I will look into how zone starts affect all of these stats in a future post), Sislo was often facing easier opposition and favorable starts, but being able to generate or take shots is how the game is won, so that’s a positive look at Sislo. Zajac, Clowe, Elias, and Bernier are slightly quickly to be involved in shots than the rest of the forwards.

This is also where it evens the playing field for Brunner and Ryder, who are certainly more apt to shoot than pass. They are only just behind those four I just mentioned, and are ahead of Loktionov, so certainly passing isn’t everything. But look at Zubrus: even playing a majority of the season with Zajac and Jagr, he still is the sixth-slowest forward to be involved with a shot. Gionta just looks bad everywhere.

Frequency of Goal Involvement

I wanted to introduce this idea of looking at which players generate a goal and combining that with the even strength goals they score during the season. It’ll be something that paints more of a picture over time with more data. But with all the frequencies of attempts and shots, I thought, what would it look like with goals?

Def_gi

Obviously Zidlicky leads the group, but the next two quickest are who? Who would you guess? Greene certainly, right? Well, no. Larsson was the next quickest to be involved in a goal, and Merrill third. Greene was fourth, tied with Fayne. Gelinas was next, followed by Salvador and a big jump to Volchenkov in last. Harrold wasn’t involved based on my criteria.

Fwd_gi

Boucher was the quickest to be involved in a goal last season at just under forty minutes. However, given his small sample size, we shouldn’t judge that too quickly. That being said, he was involved offensively for the Devils in his call-up.

Moving to the names we expect to see here: Zajac had the next best goal involvement pace at 48:39, with Elias just behind him at 48:57. Henrique was fourth at 50:23 and Jagr finished fifth at 53:10.

I’ve been talking up Steve Bernier in these pieces due to the volume he puts up and how quickly he generates it with respect to his role. So, how quickly was he involved in a goal? He was third-slowest at every 2:08:04. Why? Well, he did shoot 3.2% at even strength this season so that doesn’t help. Also, while on the ice, the Devils shot 4.7% and their PDO was 95.6 which were some of the lowest marks on the team. Bernier should be due for some regression, whether that’s by goals or assists. But, he certainly couldn’t buy a goal last season.

Now, let’s take a look and see how the Devils compared to their opposition for all of this.

Comparison with Opposition

Def_cc_si_opposition

Here we see that the Devils defensemen were quicker to be involved in both shot attempts and shots than their opposition (apart from Salvador). This is certainly affected by the team’s strong possession play throughout the season, but it isn’t as reliant on a few players as we might have thought.

Fwd_cc_si_opposition

Similar to the defense, the forwards are faster at doing both than their opposition.

So, What’d You Think?

This was the sixth article in a summer series looking at all of this data. What did you think of this data? What methods could be employed to improve the process? Give me your questions, statements, feedback so I can better steer this towards where your interest lies. Sound off below!