clock menu more-arrow no yes mobile

Filed under:

Killing Time: An Analysis of the Salvador Era of the 14-15 Devils Penalty Kill

Bryce Salvador played for the New Jersey Devils and was featured on the penalty kill for 15 games last season. This post analyzes the PK in those games in terms of how much time the team killed, who made the exits, and more in this installment of the 2014-15 Killing Time project.

Winslow Townson-USA TODAY Sports

Last week, I introduced my plans for a project I'm calling Killing Time, an analysis of the 2014-15 New Jersey Devils' penalty kill.  I am tracking each and every shorthanded situation from the previous season to identify what zone exits the Devils are making, who is making them, how the opposition enters the zone, and the time between exit and entry.  I'm calling that time between a zone exit and zone entry by the opposition as "killed time," as it's what the shorthanded team strives to do when they get the puck.  For the sake of comparison, I am splitting up last season into three "eras."  The first is what I'm calling the Bryce Salvador Era.

The team captain, Bryce Salvador, appeared in the first fifteen games of last season with significant usage on the penalty kill.  Salvador has declined as a player, he did not return to the lineup after that fifteenth game due to injury, and the PK got blown up in those games.  The goal of this analysis isn't to figure out who did or did not well on the PK or point a finger at a particular player.  However, the team's shorthanded performances were abysmal when #24 was active in the lineup.  They ranked dead last in terms of success rate. Even worse, there were only two games out of fifteen where the penalty kill did not concede any goals. I consider this to be a "worst case" group of situations; it is a good set to establish as a baseline for what I'm tracking.  We'll see in future posts whether it gets better along with the success rate.

Overall Team Summary

Here's a game-by-game breakdown of how many shorthanded situations the Devils were in, how much time they had to kill, how much they actually killed, and how many goals they conceded.  It's not pretty.

Salvador PK Era Charts

In these fifteen games, the Devils had 59 shorthanded situations and conceded 21 goals for a success rate of 64.4%.  That's terrible on it's own.  When we factor time killed into it, it's not any better looking.  Out of a total of roughly 91 and a half minutes, the Devils really only killed 35 of them. The rest was spent in the Devils' end, presumably surviving.  You can see that there have been nights where the Devils had to do a lot more PK work than others.  By the percentage killed, it does vary but has mostly been less than fifty-percent. As I've just begun this project, I can't really say what percentage or how much a "good" penalty kill should kill.  Still, spending close to two-thirds of all penalty kills in your own end defending doesn't look good.  In terms of the average number of exits per situation, the average I counted was 2.77 per situation. That does include zero-exit situations, by the way.  It provides another sort of baseline as what to look for going forward as well as a sense of what happened in these fifteen games.

Furthermore, not every penalty kill had time killed.  Nine of these shorthanded situations ended with no zone exits whatsoever.  In all nine of those instances, the Devils gave up a goal.  Often, the situation would begin with a faceoff loss, the opposition team would move it around, and score. It would be fairly quick as all nine "zero-exit" situations totaled two minutes and 42 seconds. About 42% of all goals allowed in this era happened in this way.  You can consider these to be the worst case scenarios as the Devils didn't get even get the puck out of their end once.

On the flipside, there were eleven situations where the Devils killed over half of the penalty time. I will point out that six of them were abbreviated.   Three were due to the continuation of another penalty, another three ended with goals against. Their highest percentage, 91.49%, out of their end was in a situation where the Devils were down a man for only 47 seconds. Still, there were some situations where the Devils were able to keep the opposition out of their end for at least a minute.  While I've yet to determine what's a percentage that would be considered good, I'd say 50% of the time spent outside of New Jersey's end is generally a good thing and something to strive for.  We'll see if that's true or not as the season goes on.

The first fifteen games dug a huge hole for the Devils' penalty kill.  Without spoiling too much, I will say that things will get better.  And I wouldn't necessarily say it's just because Salvador wasn't involved.  I'll get into that a little later though.

Exits, Entries, and Time Killed

In order to find out the time killed, I had to track the zone exits made by the Devils and zone entries made by the opposition.  Let's start with a summary of the types of exits the Devils made and who on the Devils made non-denied clearances.

Salvador PK Era Charts

Legend: I = exits that led to the puck going into the opposition's end, a.k.a. icings; C = exits that led to the puck going into the neutral zone, a.k.a. clears; FO = puck exits by a force out, such as a block or a deflection; X = exits by a Devil carrying the puck; P = exits by a Devil passing the puck out of the zone; UO = an exit that is unforced by a Devil, such as a missed shot caroming out of the zone or the end of a period; D = a denied clearing attempt by a Devil; GA = a goal against, which ends the penalty.

Among the exit types, the Devils were able to get more pucks into the other team's end on exits than any other.  These were mostly icings, though I did include clearances that got all the way to the goalie or the circles.  The point is that the puck was sent into the other team's end.  That's followed by a non-exit: denied clearances.  Those are attempts from a Devil that really only took a second or two (occasionally three) off the clock but never went out of the zone because the opposition stopped it or got in the way.   Coming in third are unforced outs; a mix of bad passes, missed shots, and end of periods that got the play out of New Jersey's end.  Concerningly, there were more goals against than pucks sent out into the neutral zone as well as pucks carried out, forced out, and passed out of New Jersey's end.  Again, it wasn't a good run of games for the Devils' PK.  Out of 235 events, 136 were actual exits made by a Devil.

The "Who" is by player number.  Your leader is, well, no one. Player #0 means no Devil and it is consistent with the number of unforced exits.  So the clock, missed passes and shots by the opposition, and fortunate luck had more exits than any other player.

Among actual players, Andy Greene leads the way by far with non-denied exits.  If I included denied exits, then Greene would be the clear leader. Even so, he's made more successful exits than any other defenseman, at least as twice as many as the second defenseman on the list, #7 Jon Merrill.  Forwards Patrik Elias and Adam Henrique definitely helped out more than the other forwards.  It's worth noting that Greene, Elias, and Henrique would play a significant portion of the penalty kill when active.  So it makes sense that they'd have more exits than others at their position. That said, Salvador and Mike Cammalleri also played quite a bit on the PK and didn't get as many exits. That's something to at least think about.  Most of the exits were by regulars though; Adam Larsson and Jacob Josefson would join that group eventually.  They didn't get into the picture until late in this set of fifteen games.  The group of five players with only one exit were guys who only got a few seconds on PK's anyhow.

Let's move to zone entries. How many of each type and which defenseman they tended to target.

Salvador PK Era Charts

A legend: X = opposition carried the puck into the zone; D = opposition dumped the puck into the zone; P = opposition passed the puck into the zone; E = a Devil allowed an entry by an error (e.g. turnover, pass back); FO = a faceoff in the Devils' end; NA = no entry at all; the puck either didn't leave the zone or the other team didn't get it in there.

Most of the zone entries recorded were carry-ins by the opposition.  From what I tracked, many teams have their breakout designed to have one player carry it over the line.  For some, it's because they have the space or they have a player who can force their way through.  Washington did this quite a bit against the Devils, for example, with Mike Green. For others, the puck carrier will get the puck in the zone and then immediately pass or dump it away, allowing his teammates to move forward.  Either way, carry-ins were more common against the Devils than dump-ins or passes over the line.  There were only two entries by faceoff - both penalties against the Devils - and two I marked as errors.  Both of those errors were on Greene, a turn over for one and that pass back on a faceoff that two Stars took from him and turned into a score. A bit more on that in a bit.  In total, out of 235 events, there were 144 entries.

What of the "NA" entries - the non-entries?  Well, 49 of the events were denied clearances.  Non-exits, in other words.  21 of them were goals against, which ended the penalty kill.  That accounts for 70 of the non-entries.  The remaining 21 are due to other factors like the end of the penalty while the puck is away from New Jersey's end, a penalty call on them, or the end of a period.

As far as who was targeted, Greene leads the way again.  I noticed in these fifteen games, especially in the first few, Peter DeBoer basically double-shifted Andy Greene on penalty kills. He'd start with Bryce Salvador, and then switch sides for Jon Merrill. Eventually, Damon Severson would receive a bit more time, and Adam Larsson would play quite a bit when he was in the lineup.  With more minutes and rotating between the left and right side, it makes some sense that Greene was targetted more than others.  Likewise for Salvador and Merrill, who followed Greene in minutes.  You'll notice Marek Zidlicky (#2) and Eric Gelinas (#22) on this list.  There were some kills where those two were paired near the end, presumably to spell the other defenders as the situation would end.  They were targetted a handful of times.  Mr. Nobody got two instances, based on the faceoff entries.

Now, let's get to the meat of the tracking: time.  I've identified five different types of exits - with only two instances of passes out of the zone, I've left them out - for timing purposes.  While sample sizes are definitely small here, what I've found so far is quite interesting:

Salvador PK Era Charts

After accounting for exits that led to the end of a penalty or an end of a period, each of these five types of exits were good for nearly killing close to 10% to 12.5% of a standard two-minute penalty on average.  In most of the exits I saw, they would end up lasting eleven to fifteen seconds because the opposition would collect the puck and then re-set to begin whatever breakout they want to run.  Take an icing for example.  Unless the puck's moving slowly, it only takes about three seconds to fire it at the other end of the rink.  It can take seven seconds for the opposition to recover it. Then they have to take it into the Devils' zone, which can be another five to eight seconds.  This may or may not include a line change, an extra pass on the breakout, someone moving more methodically than quickly, &c. The curious thing is that the opposition tended to do this even if the puck was just moved into the neutral zone; someone would take it back, allow players to get back into their positions for a breakout play, and then run the play.  Even if the penalty is almost up, I'd say that power play units in these fifteen games tend to pick their spots than rush an entry for a desperate attack to salvage the man advantage.

That being said, there are extemes.  You'll notice from the minimum (Min) row that there were some quick entries upon exits.  Those events weren't common. It's dependent on where players are, as the opposition players have to exit the zone to get onside and have the puck. When it does happen, it's usually because players were able to get back on-side quickly and the puck was immediately put back into play.  For example, an iced puck would be stopped by a goalie.  If they immediately fire the puck up-ice and it's retrieved by a teammate at the blueline, then that would take up half of the time an puck cleared that far would normally take.  It's not common since it's the risk - icing the puck or giving the puck back to the penalty killers - isn't necessarily as good as the reward - a quick zone entry.

In the opposite direction, some of these exits have led to extended events outside of the Devils' end of the rink. This isn't the cause of an offensive attack or a player eating up time along the boards - though, I haven't seen that yet.  It's usually the cause of a failed breakout by the opposition.  For example, the longest event outside of the Devils' zone in these fifteen games lasted 42 seconds.  Bryce Salvador iced the puck.  Dallas recovered and attempted a breakout. The Devils denied them at first. Dallas tried again and then went offside. Then there was another clearance before Dallas was able to carry the puck in on Salvador's side.  There was never a legal zone entry by the Stars, so that just added to how much time was killed.  For the Devils to extend the amount of time they kill, they would have to get a stop in the neutral zone and/or deny them at the blueline and get another clearance.  That isn't common and when it does happen, the Devils only get one or two additional stops before there is an entry.

Still, it does appear that icing the puck is a decent way of killing time provided the opposition does take it's time to recover and respond, which they have done in these fifteen games. A carry out is intuitively better - a player skating with the puck out of the zone doesn't just throw it away - but that depends on what happens after it.  If the player just ices it from the neutral zone, it's similar to an icing.  Or if they rush up ice and attempt a shot.  If they're able to take up more time or focus on denying the breakout, then the amount of time killed gets extended too.

After finishing up the second era - the remainder of DeBoer's time in New Jersey - I'd like to see if this changes.  In the meantime, let's dive a little deeper into who made the exits and entries I've recorded.

Exit Types by Player

You'll have to excuse the ugliness, but I've tried to provide a chart and a bar graph of each entry and exit to count players.  I did this to show off how stark the difference may seem from player to player. It's not pretty, but I'm trying it out. If it doesn't look good, I'll just stick to charts for future posts in this project. Again, all players are identified by player number.  Pass-outs aren't included as they only happened twice.

Salvador PK Era Charts

Greene leads the way by far in getting the puck out of his end and into the other team's.  Interestingly, Merrill came in second with ten, just ahead of three forwards: Henrique, Elias, and Cammalleri.  Salvador is somewhere in the middle.  Severson had as many as Zidlicky, which is also odd considering Severson eventually got some regular time in shorthanded situations and Zidlicky rarely appeared.

Salvador PK Era Charts

In terms of getting pucks out of the zone but not into the other team's, Greene and Salvador tied for four.  There weren't nearly as many of these clearances in these fifteen games.  The distribution of it is a bit more even  as a result, though defenders did a bit more than forwards.

Salvador PK Era Charts

As far as forcing pucks out of the zone, there weren'tmany but one would expect forwards to lead in this regard. A forward would typically apply more pressure high up and along the boards in the hopes of getting the puck out.  Henrique, Stephen Gionta, Zajac, and Cammalleri leading in this regard fits in with that.  Only one defenseman - Merrill - managed to get one of these.  I can't exactly recall, but it may have been a case of a deflected pass or something like that.

Salvador PK Era Charts

There weren't many instances of the puck being carried out.  In the few cases that it did happen, forwards commonly did it.  Greene only did it once as the lone defender.  Elias led the group ahead of Dainius Zubrus and Zajac, all regulars on the PK at the time. Martin Havlat makes a token appearance as he had the puck to start the team's first kill of the season.

Salvador PK Era Charts

Now let's get to denied clearances.  Not every denied clearance led to an extended attack, some resulted in a proper clearance just seconds later. Still, it was an attempt that didn't work out.  The breakdown by player is certainly more interesting.  Unfortunately, Greene leads in this category too.  I think it's a function of playing as much as he did. Players on the power play, particularly at the points, try to keep clearances in play.  Plus, extended work can lead to some faults in execution.

That said, Salvador follows him in second with ten.   Whereas Greene led all actual players in getting the puck out of the zone, there were only ten times Salvador did that.  He had as many denied clearances as he had successful exits.  While his exits did kill clock - two minutes and twenty-eight seconds to be exact - and at a similar average to Greene's (his exits totalled six minutes, forty-four seconds) of at least fourteen seconds.  Yet, it's telling that he failed just as much as he succeeded at getting zone exits given his significant usage on the PK.  From what I saw in tracking, Greene was able to get to more loose pucks and hustle to make plays much more than Salvador. In the captain's case, it was more of a matter of "right place, right time."  So while Greene led in denied clearances and even had some errors, Salvador's denials stick out to me in a more negative light than Greene's.   I can chalk up Greene's in part to playing so much; Salvador just wasn't so frequent or effective in clearing the puck.

In a not-too-distant third is Dainius Zubrus. As a forward, he would be higher up in the zone and theoretically have an easier time getting clearances. However, he would get into traffic along the boards and had some issues in that regard.  Like Salvador though, Zubrus having six denials to go with nine actual exits is somewhat concerning.

Entry Types by Targetted Defenseman

Let's get into zone entries by the opposition. I charted and graphed three as there were only two errors and two entries by faceoffs.

Salvador PK Era Charts

Carries happened more frequently than any other entry by the opposition. Greene was targeted more than anyone else. Again, it's a function of playing so much.  He tended to be on the right side when he was paired with Salvador. When Merrill came onto the ice, Greene would be on the left.  Some breakouts by the opposition tended to pick one side, and I'd like to think they tried to target a more fatigued Greene.  That said, opposing players did enjoy carrying it against Salvador.

Among the rest, Severson got tested a bit, possibly a function of being a rookie on the PK.  Zidlicky wasn't out there much in shorthanded situations, but opposing teams tended to go against him when he did.  Merrill and Larsson weren't tested all that much, in retrospect.

Salvador PK Era Charts

Greene got dumped-in against more than others, but Merrill nearly had as many.  Generally, that's a good thing as a team with a man advantage throwing the puck into space tends to be good for the penalty killing units that get it.  From what I saw, it was.  Salvador and Severson didn't draw as many dumps and Zidlicky only had it happen to him once.

Salvador PK Era Charts

Regarding pass-ins, or pucks passed into the zone for an entry, Salvador finally gets to lead in a category.  It's not a good one to lead into, a pass-in is similar to a carry-in.  It's a challenge to that side of the ice and Salvador drew a few more of those challenges than Greene. Opposing teams didn't go for this option though.  Ideally, a carry-in is the best type of entry and break out plays used were designed to find a spot for that to happen.  It generally worked well against the Devils.  If we're being picky, the best type of entry is the start of a power play that leads to a long, long stay in New Jersey's end - or a quick one as a result of a goal scored.  Anyway, opposing teams passed it in more against Salvador and Greene than anyone else.

Conclusions - So Far

I'm looking forward to tracking the next twenty one games.  The penalty kill was lit up for goals against in the first fifteen. It didn't happen as much in the next twenty one. I'm curious to see how much the exits and entries change with Salvador no longer returning to the lineup, Larsson getting more ice time, and others getting involved.  Greene was heavily used in the first fifteen games on penalty kills.  I don't think that will really change, as he finished the season just under four minutes per game of shorthanded ice time. I think this usage was too much at times and contributed to some of those denied clearances and other instances of faltering on the PK.

That said, he was far more effective in terms of killing the penalty than Salvador and any other defenseman at the time.  He not only attempted more exits, but he achieved many more exits.  He did this as well with Merrill on his side as opposed to the similarly-used Salvador.   From the standpoint of actually killing clock, #6 did more than most, including #24.  It belies the notion that Salvador is somehow better on the penalty kill. While it's true his lack of mobility would be somewhat sheltered and he wouldn't have to try to move the puck to start an offensive play, he just didn't get to a lot of pucks for clearances.  He may have helped make clearances possible, but he wasn't one making the important play to kill time.  Greene was.  So while Greene got denied more and had some errors, I came away from these fifteen games worrying about his back since he carried so much of the PK effort on it.

From a time standpoint, again, it's too early to say whether the percentage killed is what should be expected or whether it was worse than a regular penalty kill.  That's why I'm looking forward to the next era, which should be done in the next few weeks.  At a minimum, I have the expectation that most clearances will take ten to thirteen seconds off the clock.  We'll see if that holds true or not.  If it is, given an average of 2-3 exits per situation, perhaps it should be expected that the Devils will kill 30-45 seconds of a common two-minute penalty.  But that's partially why I'm doing this: to see how much of the penalty the Devils killed last season.

In the meantime, I'm going to track the next 21 games.  I want to know what you think. What have you learned from the results of the fifteen games I've tracked so far?  Are you surprised to have seen the Devils kill less than 40% of time of all penalties in these games? Do you come out of this post thinking more or less of Andy Greene doing so much?  Is there something else you want to see in these posts based on what I'm tracking?  Please leave your answers and other thoughts about the Salvador era of the Devils' penalty kill from last season in the comments.  Again, the next post will be up in a few weeks, and I promise you, the PK does get better from a success standpoint. Thank you for reading.