Killing Time Pt. 4: An Analysis of the 14-15 Devils Penalty Kill Under the Three Coaches

After a month, I give you today the fourth full installment of the Killing Time project.  For the unaware, this is the analysis of the New Jersey Devils' penalty kill, situation by situation, where I track how much time the Devils kept the opposition out of the zone, who made the zone exit (or who got denied by it), who got targeted on the entry, and what kind of zone exit was made.  In Part 1, I explained my intentions and goals for this project. In Part 2, I released the results of the first era of last year's penalty kill, the Bryce Salvador era - named for the captain who was active and heavily featured on the PK at the time. In Part 3, I released the results of the second era of last season's penalty kill, the Peter DeBoer era - named for the head coach who was still behind the bench without Salvador and until his firing in late December.  While tracking the remainder of the season, I put up this interlude to the project detailing that not much changed tactically before and after the coaching change. This is Part 4, the penultimate post of this project.  It has the results from what I've called the Multiple Coach or MultiCoach era - named for the three head coaches behind the bench.  While they didn't necessarily run the PK, they did decide who would be on the ice and oversaw how the team confirmed their bottom-ten finish in the league.

While how I broke it up led to three differently sized sets of games - 15 for the Salvador Era, 21 for the DeBoer Era, and 46 for the MultiCoach Era - it did provide three even situations. The first era was the worst case scenario, when the PK got wrecked. The second era was the best case scenario, when the PK success rate exceeded 90%.  This third era lies somewhere between brilliant and brutal.  From December 27, 2014 through to the end of the season, the Devils' penalty kill had a success rate of 82.6%.  Thanks to NHL.com's new filters for their stats, I can tell you that in that time span, the Devils' PK ranked 13th in terms of success rate.  In other words, they were good but not exceptional.   Over the whole 82 game season, the Devils finished at 80.6% at 21st overall. That's a testament to how terrible they were in the first five to six weeks of the season and how difficult it is to overcome such a deficit.   All the same, these 46 games should be seen as closer to what the PK could perform when they're not ice cold or blazing hot.

Overall Team Summary

Let's begin with a game-by-game breakdown of how the PK performed with respect to how much time they kept their opponents out of their zone.

Game Date	# of PK's in Game	Total PK Time in Game	Killed Time Total	% Killed	# PPGA
12/27/14	3	04:46	01:49	38.11%	1
12/29/14	3	06:00	02:23	39.72%	0
12/31/14	4	08:00	02:43	33.96%	0
01/02/15	4	07:09	01:59	27.74%	1
01/03/15	2	03:28	02:08	61.54%	1
01/06/15	2	03:31	01:31	43.13%	1
01/08/15	3	03:14	01:09	35.57%	1
01/09/15	4	06:00	02:07	35.28%	0
01/14/15	4	08:00	04:05	51.04%	0
01/16/15	3	04:45	01:52	39.30%	1
01/19/15	4	07:48	02:06	26.92%	0
01/28/15	5	10:00	04:02	40.33%	0
01/30/15	3	04:18	02:01	46.90%	1
01/31/15	2	03:01	01:20	44.20%	0
02/03/15	1	02:00	01:04	53.33%	0
02/06/15	2	03:30	01:12	34.29%	1
02/07/15	3	04:05	00:58	23.67%	3
02/09/15	3	05:15	01:48	34.29%	1
02/13/15	4	07:31	02:39	35.25%	1
02/14/15	2	04:00	01:19	32.92%	0
02/17/15	3	06:00	02:42	45.00%	0
02/20/15	3	06:00	02:21	39.17%	0
02/21/15	2	04:00	01:52	46.67%	0
02/23/15	4	06:17	02:52	45.62%	0
02/25/15	2	02:35	01:02	40.00%	1
02/27/15	2	04:00	01:51	46.25%	0
02/28/15	3	06:00	03:16	54.44%	0
03/03/15	3	06:00	01:48	30.00%	0
03/06/15	2	02:04	01:13	58.87%	0
03/08/15	3	05:43	01:24	24.49%	0
03/10/15	4	07:19	02:34	35.08%	1
03/12/15	3	04:27	01:49	40.82%	0
03/14/15	1	02:00	00:42	35.00%	0
03/17/15	0	00:00	00:00		0
03/20/15	4	06:28	03:08	48.45%	1
03/21/15	1	02:00	00:54	45.00%	0
03/23/15	3	05:32	01:32	27.71%	1
03/26/15	1	02:00	00:55	45.83%	0
03/28/15	3	06:00	02:44	45.56%	0
03/29/15	1	02:00	00:41	34.17%	0
03/31/15	4	07:58	03:51	48.33%	1
04/03/15	6	12:34	05:34	44.30%	1
04/04/15	1	00:50	00:05	10.00%	0
04/07/15	7	10:40	03:59	37.34%	1
04/09/15	8	13:08	05:29	41.75%	3
04/11/15	3	06:05	02:29	40.82%	1
Total	138	244m1s	96m59s	39.74%	24

In the first part, the Devils kept their opponents out of their end 38.47% of the time.  In the second part, it was 39.9% of the time.  In these 46 games I recently finished tracking, the team kept their opponents 39.74% of all shorthanded ice time.  There were only five games where the Devils spent 50% of the total shorthanded ice time out of their own end of the rink.  This further speaks to my working theory that an effective penalty kill does manage to keep their opposition out of their own end, but it's more important that they defend their own end.   Given the inherent match-up issue on penalty kills, it shouldn't be a surprise that the opposition spends a majority of their time on offense. Therefore, an emphasis should be placed

That does not mean being able to take time off with zone exits are a bad thing or not important.  It just can't be the only thing to judge a PK anymore than it is to judge them by whether they give up a goal or not.  I'm getting into "Lessons Learned" territory, which I'd like to save for a final post next week. But after tracking 82 games, it's worth highlighting that now.

In any case, with a wider range of games, there were just more occurrences of everything.  More zone exits. The average of exits per situation rose to 3.24 in this era.  More zone entries. That also rose as, well, there were more exits.  More zero exit situations.  There were ten of those.  Surprisingly, only five of those ended with goals against. There were two actual full two-minute penalty kill with no exits whatsoever and the other three were truncated situations.  The total amount of zero-exit situations that I counted was six minutes and 43 seconds; a definite bump up from the DeBoer era that only had one.

There were also more personnel changes.  The defense saw Seth Helgeson leave as #25 and return for one game at #33 at the end of the season.  Raman Hrabarenka (#34) make a cameo appearance.  Adam Larsson (#5) became a mainstay with Andy Greene (#6).  You'll see how much of a mainstay that is.  You'll also notice what you don't see. Damon Severson (#28) made no exits and was essentially kept off of the penalty kill when he became able to play again.  Jon Merrill saw his usage cut, more in favor of Mark Fraser (#32) and even Peter Harrold (#10).  Up front, Stephen Gionta (#11) became a big PK'er when he came back; Tim Sestito (#12) and Tuomo Ruutu (#15) faded away; and you'd be surprised who led all of the forwards in shorthanded ice time.  Some of this is natural in a set of 46 games, but it was a sort of a surprise to see less of Merrill and a lot less of Severson.  Likewise, it was not at all a surprise that 5-6 was leaned on so heavily.

Exits

Let's start with the types of zone exits.  There were more of everything:

Exit Types	Count	% Total
I	223	38.92%
D	95	16.58%
UO	73	12.74%
C	63	10.99%
FO	42	7.33%
X	40	6.98%
GA	24	4.19%
P	11	1.92%
NA	2	0.35%

Once again, (I)cings were the most common type of zone exit.  Shorthanded teams are allowed to clear it the length of the rink with no penalty and so the Devils have done that.  Its occurrence dwarfs the next type of exit, which isn't an exit at all. It's (D)enials, the count of a Devil clearly attempting to clear the puck but getting stopped by the opposition or their own teammate from doing so.  Unforced exits (UO) return to a more prominent position, although they do include shorthanded situations started outside of the zone and when new periods begin.  (C)lears are those exits that just get beyond the blueline but do not go through the neutral zone. They happened with some regularity. Force outs (FO) are exits caused by deflections, poke checks, re-directions, blocks, and all other miscellaneous plays by a Devil that get an exit but weren't from a possession.  They were followed closely by Devils carrying the puck out (X). Curiously, both were similar in terms of counts but didn't happen nearly as often as the other exits. Goals against (GA) obviously represent an exit as they ended the penalty kill; unfortunately they did not finish last.  Passes out of the zone (P) brought up the rear along with two exits I had to call (NA) as there were two situations with no exits whatsoever.  Overall, the breakdown in exits is similar to prior eras, though unforced exits rose dramatically compared to the last era I looked at for Killing Time.

Since unforced exits rose and there can't be any denials for them, no Devil - represented by #0 - led in exits during this era.

Who Exited	Count	% Total	Denials	SHTOI (min)	Exit/Min	Exit/2 min	Exit/60 min
0	73	16.19%	0	0	NA	NA	NA
6	65	14.41%	21	170.12	0.38	0.76	22.92
5	49	10.86%	20	151.74	0.32	0.65	19.38
19	36	7.98%	7	77.31	0.47	0.93	27.94
26	36	7.98%	6	56.66	0.64	1.27	38.12
14	31	6.87%	4	65.31	0.47	0.95	28.48
16	29	6.43%	6	82.23	0.35	0.71	21.16
11	27	5.99%	0	57.42	0.47	0.94	28.21
8	25	5.54%	2	51.48	0.49	0.97	29.14
23	20	4.43%	4	41.23	0.49	0.97	29.11
7	18	3.99%	9	52.89	0.34	0.68	20.42
32	16	3.55%	10	57.64	0.28	0.56	16.66
10	10	2.22%	3	34.45	0.29	0.58	17.42
12	9	2.00%	1	22.67	0.40	0.79	23.82
15	2	0.44%	0	10.43	0.19	0.38	11.51
25	2	0.44%	1	4.5	0.44	0.89	26.67
33	1	0.22%	0	2.92	0.34	0.68	20.55
34	1	0.22%	0	1.32	0.76	1.52	45.45
18	1	0.22%	0	0.27	3.70	7.41	222.22

There were a total of 452 exits in these 46 games. One was made by Cory Schneider, so this chart is for the 451 made by a Devils skater or no skater at all.  Following abeyance, Andy Greene and Adam Larsson lead the rest of the way.  They made the most exits. They got denied the most, too.  The reason for both is simple: they played the vast majority of shorthanded ice time in this range of games.  No Devil comes close to them. The only reason why Larsson didn't play as much as Greene was because Larsson took more penalties than Greene.  Even so, over 150 minutes each on the PK is a stunning amount.  What's more stunning is their rates of clearances. Only Jon Merrill and the seven or so minutes of Seth Helgeson comes close to their rates.  Harrold and Fraser were used more sparingly, but their lowered clearance rates provides evidence as to why that is.  While Larsson and Greene were surely pinned back more in their extended time, they were able to get more attempts to clear the puck than the other two.  In retrospect, one may wonder why Merrill didn't get more shorthanded ice time.

Among the forwards, your leader in shorthanded ice time was Jacob Josefson (#16) for this set of games.  He was around the middle in terms of total exits made.  Perhaps it is a function of so much ice time?  Then again, Travis Zajac played five fewer minutes and was able to make more clearances in terms of gross and rate.  Patrik Elias and Adam Henrique did they same in less time, while Stephen Gionta (who was out hurt for a portion of this era), Dainius Zubrus, and Mike Cammalleri had superior rates of making exits.  Even Tim Sestito had a better rate in his limited usage.  One can't say the three coaches didn't give Josefson a chance on special teams. Though, one may say he could have been more effective with less time given.

Time Killed by Exit

While on the subject of entries, I once again split up the amount of time each exit killed by the type of exit. I removed instances where the penalty or the period ended to keep the exits on a more level playing field.  Since there were at least ten exits by passes, I included that type for this first time in this project:

	Time Killed by Exit Type in Seconds During MultiCoach PK Era – End of Period/End of Penalties Removed
Type of Exit	Icing	Clear	Force Outs	Carry Outs	Unforced Outs	Pass Outs
# of Exits	189	58	36	35	29	11
Min	2	1	1	5	3	4
Max	48	32	42	41	23	33
Average	14.39	11.07	13.75	16.54	11.72	15.22
Std. Dev.	5.52	6.4	8.45	6.91	5.52	9.11
Time Killed (in Min)	45.3	10.7	8.25	9.65	14.36	2.28

A quick word about the minimum time taken off the clock by the exits.  Those one and two second minimums were cases where the puck went out of play and so play returned to the Devils' end.  Only a second or two is taken off the clock when it goes into the bench.  Removing those instances would leave their minimum amount of time removed to be four to five seconds, consistent with carry outs, unforced outs, and pass outs.  Also, look at the maximums.  There were only a few cases where an exit led to more than a half of a minute being killed.  In all of them, there were extra events outside of the zone that led to more time being taken off, such as a breakout being denied, a forecheck, or extended possession.

While icings led the way in terms of the number of exits, exits by skating the puck out took off the most time on average. This makes some intuitive sense as the puck carrier does have options. They can skate it out for a soft dump. They can take it up ice and just maintain possession. They can go for a shot, recover the rebound, and pass it around.  Likewise with passing the puck out of the zone; it creates a possession and with possession, there are more possibilities than just throwing it away.  That said, those opportunities are limited in a shorthanded situation.  It's more often than not that the Devil has to just fire it away, and so icings were more common.  They took off an average amount of 14.39 seconds, which isn't too bad at all.  Much better than just clearances into the neutral zone or forcing the puck out, as both could (and did) lead to some quick entries by the opposition.  Overall, an exit took off an average of at least 11 seconds in these 46 games.  That's nearly a tenth of the common two-minute minor penalty, so getting an exit at all in a pinch is more important than focusing on a certain type.

Exit Types by Player

Since icings were the most common type of exit, let's see who led the way in successfully making them.

Who "Iced" the Puck	Count	% Total	Icing/Min
6	51	22.87%	0.30
5	42	18.83%	0.28
26	17	7.62%	0.30
19	16	7.17%	0.21
11	14	6.28%	0.24
14	14	6.28%	0.21
7	13	5.83%	0.25
8	12	5.38%	0.23
23	11	4.93%	0.27
16	11	4.93%	0.13
32	8	3.59%	0.14
10	6	2.69%	0.17
12	4	1.79%	0.18
15	2	0.90%	0.19
33	1	0.45%	0.34
25	1	0.45%	0.22

Andy Greene remained your icing king through these 46 games.  Larsson finished in a somewhat closer second place and the rest of the team made up the numbers.  Some of their rates, such as Patrik Elias', look real good but that's also a function of playing far fewer minutes than #5 and #6.  That the two defensemen played over 150 shorthanded minutes and still managed the best rates (Seth Helgeson's one game wearing #33 aside) speaks to how they were committed to getting the puck out to the other end of the rink.  Incidentally, plenty of these came off faceoffs that the Devils won.  In retrospect, I wish I noted those instances just for clarification's sake.

Who "Cleared" the Puck	Count	% Total	Clear/Min
11	7	11.11%	0.12
16	7	11.11%	0.09
26	6	9.52%	0.11
5	6	9.52%	0.04
6	6	9.52%	0.04
23	5	7.94%	0.12
14	5	7.94%	0.08
10	4	6.35%	0.12
7	4	6.35%	0.08
32	4	6.35%	0.07
12	3	4.76%	0.13
8	2	3.17%	0.04
19	2	3.17%	0.03
25	1	1.59%	0.22
35	1	1.59%

In terms of just getting the puck beyond the Devils' blueline, it was a much closer competition.  Stephen Gionta and Jacob Josefon tied for first with Elias, Greene, and Larsson right behind them.  Some of the clearances would've been icings if it wasn't for a deflection or a touch by another player slowing the puck down enough for the opposition to retrieve it in the middle.  These are not ideal exits compared to icings if only for how much time they can take off the clock, but it is better than none at all.  Note that Schneider made one of these to bring up the end of the list.

Who "Carried Out" the Puck	Count	% Total	Carry/Min
19	9	22.50%	0.12
8	5	12.50%	0.10
16	5	12.50%	0.06
26	4	10.00%	0.07
11	4	10.00%	0.07
14	4	10.00%	0.06
6	3	7.50%	0.02
23	2	5.00%	0.05
18	1	2.50%	3.70
34	1	2.50%	0.76
12	1	2.50%	0.04
5	1	2.50%	0.01

If Gionta was healthy and active during all 46 games, then I wonder whether he would've led the Devils in exits by carrying the puck out.  He did it quite a lot and more often than the rest under DeBoer.  That didn't happen nearly as much.  Instead, Zajac managed to have the most, followed by Zubrus and Josefson.  #19 made a point of it to move it out himself for an array of options.  Some were attacks, some led to shots, and others just were there to take time off the clock.  This type of exit features Steve Bernier, who had one at the end.  His rate looks super high since he only played about 26 seconds of shorthanded ice time.  Likewise, Hrabarenka's carry out rate is boosted by his limited ice time.

Who "Forced Out" the Puck	Count	% Total	Force/Min
19	8	19.05%	0.10
26	7	16.67%	0.12
14	7	16.67%	0.11
16	6	14.29%	0.07
32	4	9.52%	0.07
8	3	7.14%	0.06
6	3	7.14%	0.02
11	2	4.76%	0.03
23	1	2.38%	0.02
7	1	2.38%	0.02

Zajac did not just carry the puck out more than others, he also forced the puck out with stick checks, his body, and so forth.  He just finished ahead of Elias, Henrique, and Josefson.  Forwards should be leading this category as they are more likely to pressure puck holders on the power play.  With the Devils' PK switching into a wedge plus one from a box, that "plus one" is usually a forward. That allows them to get up and make an exit from time to time.

Who "Passed Out" the Puck	Count	% Total	Pass/Min
8	3	27.27%	0.06
26	2	18.18%	0.04
6	2	18.18%	0.01
12	1	9.09%	0.04
23	1	9.09%	0.02
14	1	9.09%	0.02
19	1	9.09%	0.01

Since I included pass outs, here's the table for that.  Not much to say other than that I'll bet you're a little surprised Zubrus is the leader here. Or maybe not, as Elias and Greene were just one completed pass out of the zone shy of tying him.  In general, passing out of the zone doesn't happen a lot as it requires a Devil to be out of the zone ahead of the puck. That's generally not a good idea in a shorthanded situation, and so the opportunity rarely presents itself.

Who Was "Denied" a Clear	Count	% Total	Denial/Min
6	21	22.11%	0.12
5	20	21.05%	0.13
32	10	10.53%	0.17
7	9	9.47%	0.17
19	7	7.37%	0.09
26	6	6.32%	0.11
16	6	6.32%	0.07
23	4	4.21%	0.10
14	4	4.21%	0.06
10	3	3.16%	0.09
8	2	2.11%	0.04
25	1	1.05%	0.22
12	1	1.05%	0.04
35	1	1.05%	#N/A

Lastly, to close out the Exits section, here's the table of who got denied on an attempted exit.  Greene takes the top spot, but Larsson is just behind him.  In addition to being the two skaters who cleared more pucks than any other Devil, they also got denied more than the rest.  It is a function of being in a position to make a clear for so many minutes.  Denials are bound to happen with such exposure to each situation.  The curious part of this list are who's behind them: Mark Fraser and Jon Merill.  Both got denied more than any non-Larsson and non-Greene defenseman, but recall they played a lot of less than both and they made far fewer exits. In fact, Merrill's denials are half of his exits and Fraser's denials are more than half of his exits.  This strongly suggests that while Larsson and Greene got denied more, they were more active and efficient in their attempts than Merrill and Fraser. It may explain why they were kept on secondary units unless necessary.

Entries

Entry Types	Count	% Total
X	223	38.92%
P	89	15.53%
D	72	12.57%
E	5	0.87%
FO	5	0.87%
NA	179	31.24%

Opposing players still preferred to carry it in (X) against the Devils. I saw an uptick of entries by passes (P) in this era. Some of those carry-ins were really passes upon a second look.  It's difficult as it is common for an open man at the blueline to receive the puck to make the entry. Sometimes that's beyond the blueline, but other times it's right before it. The former is technically a passing entry, the latter is technically a carrying entry.  All the same, both are preferable for a power play than a dump-in (D), which still happened here and there.  With 46 games, there were bound to be a few more errors (E) by the Devils in their own end that created the entry and faceoffs (FO). Neither happened so much to be worth noting in a bigger picture.  No entries (NA) were again a combination of goals against and the ends of both penalties and periods.  Those happened more often than everything but carry-ins, which again speaks to how prevalent they were. It makes intuitive sense. A power play has one extra man, that man should be able to get open or open the defense up enough to make space for a carrier to get in the zone.

Who Targeted	Count	% Total
5	112	19.55%
6	104	18.15%
7	69	12.04%
32	60	10.47%
10	31	5.41%
25	6	1.05%
2	2	0.35%
28	2	0.35%
33	2	0.35%
34	1	0.17%
NA	184	32.11%

Your eyes are not playing tricks on you, there were more "NAs" targeted than "NA" entries. With the five faceoff entries, I had to assign the targeting to no one in particular. The numbers do add up.  In any case, Larsson was targeted more than any other defenseman.  Greene got a lot too, but Larsson just had a few more.  Over this set, I've noticed that some teams intentionally breakout towards a side of the rink regardless of who is on that side.  There have been situations where, say, four out of five entries were on the right side. Even if the left side was guarded by the slow and not-too-effective Mark Fraser, the opposition still went right.  It helped Larsson take the lead despite playing about twenty minutes less than Greene.  It helped Merrill take third despite his lower amount of ice time being paired with defenders we may consider worse.  It helped Harrold not get targeted as much.  With entries, there were a few cameos for Marek Zidlicky (#2) and Severson; those were instances just before the penalty ended.

Let's break down these entries by type. First, the carry-ins.

Targeted for Carries	Count	% Total
5	65	29.15%
6	60	26.91%
7	40	17.94%
32	30	13.45%
10	20	8.97%
25	3	1.35%
2	2	0.90%
33	2	0.90%
34	1	0.45%

Again, Larsson "finished" ahead of Greene. That Merrill also "finished" further ahead of Fraser and Harrold speaks to that earlier point.

Targeted for Pass-ins	Count	% Total
5	28	31.46%
6	24	26.97%
32	15	16.85%
7	14	15.73%
10	6	6.74%
25	1	1.12%
28	1	1.12%

As for pass-ins, Larsson just came ahead of Greene again.  Fraser got targeted just a bit more than Merrill on passes, which may speak to come kind of equilibrium among this type of entry.  The idea of the pass-in is to get the puck to someone who can play into that space.  That may go against preferences to one side or the other on a breakout.

Targeted for Dump-ins	Count	% Total
6	19	26.39%
5	16	22.22%
7	15	20.83%
32	14	19.44%
10	5	6.94%
25	2	2.78%
28	1	1.39%

Dump-ins are curiously even among the top four. Greene came in first, but Merrill and Fraser finished closely behind. It could be that as a defenders on secondary units, whether it's Merrill and Harrold, Merrill and Fraser, Fraser and Harrold, or something else, they will go up against secondary power play units. With lesser opposing players, a dump-in may be attempted just to fire the puck in, more often than not.  I would say it'd be a tactical decision if someone slow like Fraser was targeted way more than the rest on these.  However, that didn't happen in this set of games - or in the prior sets.

Conclusions So Far

As I'm going to make the final post a summary of all data and a "Lessons Learned," I'm going to hold back on some of the conclusions I came up with in these 46 games.  I think some of them are fairly obvious with respect to personnel, how different exits took time off the clock, and how much time was taken off the clock.  Again, I recognize the disparity in the number of games between all three eras.  That said, I am pleased that my breaking points allowed for a worst case, best case, and somewhere-in-the-middle case scenario to be looked at separately.   These 46 games certainly represent a somewhere-in-the-middle scenario and it's not a bad one.  It's success rate was better than the league median for its time frame.

There weren't many situations where I just watched with my mouth agape in shock.  Despite my lamenting at their sight, Harrold and Fraser weren't so abysmal on the PK.  They had their moments of awfulness, but so did everyone else at some point or another.  Guys who have been declining like Zubrus and Elias didn't look so declining. That all said, the PK rode Greene and Larsson hard.  Their presence was often missed when not available on the ice. Likewise, Schneider and Keith Kinkaid had plenty of moments where they needed to be the team's best penalty killer. How much of that is just a good power play as opposed to an ineffective penalty kill, I cannot say.  What I can point to is how many exits they made, how much time that took off the clock, and how the opposition entered the Devils' end to renew their attack.

I'll have a fuller summary and better conclusions next week.  In the meantime, please let me know what you think of this penultimate post of the Killing Time project.  What have you learned from this?  What surprised you and what didn't surprise you from the results of 46 games of penalty kills?  Please leave your answers in the comments.  If you appreciate this post and others we do here at this site-yet-to-be-renamed, then please share it with others you know. As always, thank you for reading.