With the Devils taking steps towards rounding out the roster this month, let’s look at some negative trends from the previous season that should be addressed in September.
With the New Jersey Devils roster largely set for the 2025-26 season and everyone putting their best lineup prediction out on the internet, I have been thinking about how Sheldon Keefe can improve usage in his second year with the team. Despite the anxiety felt by Devils fans — that they have not added enough to the roster, and that too many of the players remain the same — there are very few moves left that I think Tom Fitzgerald can make this offseason. For the most part, this is the group that Keefe has to work with in the fall.
Penalty Kill Deployment
By the numbers, the worst penalty killers on the New Jersey Devils were among the players most utilized by Sheldon Keefe and Ryan McGill. Now that McGill has been replaced by Brad Shaw, most recently of the Philadelphia Flyers, some of these issues may resolve themselves in terms of structure and approach. However, I think it is worth considering that the Devils have not made their best choices in terms of usage when disadvantaged. We have seen the ill effects of the penalty kill at some particularly bad moments, despite overall positive numbers, such as in Game Five against the Carolina Hurricanes. But we have also seen some fantastic stretches where the defense and Markstrom seem impenetrable. So, how can the Devils eliminate those bad moments, solidifying themselves as one of the top penalty kills in the NHL?
The PK Performance Disparity
Brenden Dillon ranked first on the team in penalty killing time on ice with 215:34 in 2024-25. Johnny Kovacevic followed close behind at 201:55. In total, they played 156 of those minutes together. In those minutes, the Devils scored two shorthanded goals and allowed 24 power play goals against. This was not too far off from their expected numbers, as the team allowed over 54 shots per 60 minutes with those two on the ice on the penalty kill. The 9.23 goals allowed per 60 penalty killing minutes with this duo defending is almost double the rate the Devils gave up with other penalty killing defensemen on the ice.
Nico Hischier and Dawson Mercer are a similar story. Mercer led all forwards with 161:37 of penalty killing ice time, with Nico in second at 154:42. Nico had the worst on-ice goals against per 60 penalty killing minutes at 11.64, with Dawson Mercer second worst among all Devils at 8.91. While you might expect these two to chip in some offense on the penalty kill, that also did not happen. As a duo, they underperformed their expected goals on both ends, scoring one shorthanded goal compared to 2.54 expected, while the team allowed 21 goals compared to 19.1 expected goals against.
But Dawson Mercer was on the ice for three goals against on his own (36:57), while Nico Hischier was on the ice for a whopping nine goals in 30 penalty killing minutes played without Mercer, with the Devils scoring one shorthanded goal in that time. Interestingly, Mercer overperformed his 6.12 expected goals against, while Nico’s on ice results were a massively underperformance for the team’s 5.64 expected goals against in his minutes.
So, just how bad is it? Hischier and Mercer account for 191:38 of the Devils’ penalty killing time. During those minutes, the Devils scored two shorthanded goals and allowed 33 power play goals against. The Devils killed 198:52 of penalty time without Hischier or Mercer on the ice. In those minutes, the Devils scored five shorthanded goals and allowed just seven power play goals against.
Who did well for the Devils on the penalty kill, though? Brett Pesce only saw 11 goals against while the team scored four shorthanded in his 157:12 of penalty killing time. He should probably be the number one penalty killing defenseman over Dillon or Kovacevic (even after Kovacevic’s return from injury), despite Kovacevic’s positive results away from Dillon. Him and Jonas Siegenthaler are an excellent defensive duo and should get the majority of minutes in 2025-26. Jesper Bratt also saw four shorthanded goals scored to 10 power play goals against in 125:27 of penalty killing time. The question is: who should Bratt play most with?
Who Plays Center?
In 17 minutes together, Hischier and Bratt had disastrous results. They allowed six power play goals and scored one shorthanded, giving up over a shot per minute. With any other center, Bratt was on the ice for the Devils giving up just 0.59 shots against per minute on the penalty kill. Nico Hischier is a first-line center. That is not in doubt. But what makes him such a good player is how he wins matchups at five-on-five and how effective he is on the power play (and, increasingly, how high-percentage his shot is). Penalty killing defense is a different beast, and Hischier has not figured that one out yet. See HockeyViz’s visualization of Hischier’s man-disadvantage effect below. In those minutes, the Devils give up a ton more shots at the point and right in front of the net, while the left wing one-timer is wide-open, just above the dot. That’s a recipe for penalty killing disasters.
HockeyViz
And see how that plays out in real terms, with Hischier struggling to shut down the middle of the ice on the penalty kill:
Many of the minutes Bratt played on the kill were with Erik Haula. Haula was incredible in his penalty killing minutes, with the team averaging under 2.5 goals against per 60 minutes regardless of who was on his wing. Had Tom Fitzgerald traded someone else to create cap room, that would have been a perfect top penalty killing unit. But the Devils will have to look elsewhere. The next-best option is, unintuitively for many people, Jack Hughes. In all of his 51:25 of penalty killing ice time, the Devils outscored opponents 2-1. Is that sustainable? Almost certainly not to that extent, but Hughes had a whopping 30.03 xGF%, showing that he might actually be severely underplayed on the penalty kill.
Increasing Depth Utilization on the Penalty Kill
Otherwise, the Devils would have to hope that Cody Glass can become a full-time penalty killer. In his career, Glass has only played 98:34 of total time on the kill, with his team allowing seven goals on 12.92 expected goals against. That comes out to about 7.86 expected goals against and 4.26 actual goals against per 60 minutes of penalty killing time in his career, albeit in a tiny relative sample. If those rates held up over a full season, Cody Glass would be the third-best penalty killing forward on the Devils, compared to those with over 50 such minutes played in the 2024-25 season (Hughes, Haula, Bratt, Mercer, Hischier).
Connor Brown’s arrival will be a huge aid. Brown’s career low usage on the penalty kill was 124:52 in the 2023-24 season, and he has mostly had good results, with just 2019-20 and 2024-25 sticking out with higher goal against numbers. His expected numbers have always been solid. While Brown has not been able to achieve the same level of shorthanded threat, post-recovery, that he brought to the table in Ottawa, he should still be able to take a lot of the minutes that the team would have used Haula for last season. If you are hoping that Juho Lammikko can also be an option, though, you might want to hope that he learned a few tricks back in Europe. He was very bad for Florida and Vancouver in considerable minutes.
HockeyViz
Another option for improving forward usage is utilizing other players who played a small, but notable amount of time on the penalty kill. Two such forwards remain in the 25-50 minute usage category: Stefan Noesen and Ondrej Palat. In Palat’s 28:25 of penalty killing ice time, the Devils scored twice and allowed zero goals. In Noesen’s 38:37 of penalty killing ice time, the Devils allowed zero goals. Given the consternation about Palat’s value to the team, one apparent way to revitalize him would be to use him in these situations more often. In his earlier days, Palat spent a lot of time on Tampa Bay’s penalty kill, logging over 100 minutes three times. In his career, Palat has been on the ice for 18 shorthanded goals and 75 power play goals against in 761:48 of total penalty killing ice time. That comes out to 1.42 shorthanded goals for and 5.9 power play goals against per 60 minutes of ice time Palat has played on the kill in his career. He might be fading at five-on-five, but players often retain their special teams ability as they age. On the other hand, Noesen is effective in all situations, but can probably pick up more penalty killing minutes if he finds himself off the first power play in 2025-26.
What is the importance here? If the Devils’ top penalty killers remain Hischier, Bratt, and Mercer, and two of those players remain ineffective penalty killers, the Devils are wasting their minutes. Over the course of a game or season, how much more fatigued is Nico Hischier because he takes top penalty killing matchups? Taking him out of that role might hurt his chances of winning a Selke award, but there are either two roads there. Either Nico improves as a penalty killer and actually puts himself into that conversation, or the team pulls back on those minutes to put him in more offensive situations, where he thrives most. Hockey media talks all the time about how important good fourth lines are to take the burden off top players, giving them time to rest. The Devils can get somewhere in that regard by relying more on their supporting cast to kill penalties.
Unlikely Blueliners
While I have advocated for Siegenthaler and Pesce taking the top unit matchups away from Dillon and Kovacevic, even after the latter’s return from injury, I think the defensive use share can be shaken up even more. Devils fans might not think of Dougie Hamilton, Simon Nemec, or Luke Hughes as penalty killing options, but at least one of them will have to play large minutes in the absence of Kovacevic. My take, though, is that all three should.
Let’s start with Dougie. You might be surprised to find out that the coach who believed in Hamilton most as a penalty killer was Rod Brind’Amour, who played him for over 100 minutes in 2019-20, and was on track to do so again had the 2020-21 season not been shortened. In Carolina, Dougie was on the ice for seven shorthanded goals and 25 power play goals against in 209:05 of penalty killing ice time, and topped out at a sky-high 37.73 xGF% in 2019-20, when he had his most usage. That was 2.01 goals for and 7.17 goals against per 60 minutes of kill. In a mere 20 minutes this year, Dougie had a 31.43 xGF% on the penalty kill. Why can Hamilton work there? The one and only true concern about Dougie Hamilton’s defense is his speed, which NHL EDGE’s tracking data seems to disagree with, anyway. But even if you do not believe that data, penalty killers do not defend a constant rush. They have to defend at a near-standstill, and not many players have a better reach than 6’7” Dougie Hamilton.
This is the same reason I believe Brenden Dillon is not suited to be a top penalty killer. Dillon has had negative relative impacts on the penalty kill for three consecutive seasons, and I think it has to do with his style of defense. His best plays are at five-on-five, when he has the freedom to step up and blow up forwards on the rush. On the penalty kill, he often ends up trying to defend plays by the net by dropping to one knee and blocking the ice with his stick, but he does not take net-area space away in an effective manner. As a result, shots against explode in frequency right in front of Markstrom and Allen when Dillon is killing penalties. This is supported by HockeyViz’s shot tracking data, which has regarded Dillon as a negative penalty killer across his entire career, save the 2020-21 season.
Enter Luke Hughes. Hughes is still learning how to defend at even strength, but the limited returns on his penalty killing have been positive. In just over 27 minutes on the penalty kill this season, Hughes saw an increased number of slot shots, but a strong reduction of left-side crease-area shots. In the shot rate map from HockeyViz below, you can see that the increase in shots against when Luke Hughes is on the ice comes from sharp-angle chances below the right-wing faceoff dot in addition to above the right-wing faceoff circle, near the point. Otherwise, it seems that Hughes’s defensive partners on the penalty kill were the ones allowing the majority of dangerous net-area chances.
HockeyViz
I can hear the sample size complaints already, but these are just the early returns for a player who has not been utilized in this role much in his young career. It does seem that Luke worked very well with Brett Pesce on the kill, but he struggled mightily in a 6:54 lefty-lefty sample with Brian Dumoulin, who is no longer on the team. Those minutes with Dumoulin account for two of the three power play goals against with Luke on the ice. I believe the third such power play goal against was actually in a five-on-five empty net situation, with Luke as the only defenseman on the ice, as there was no other registered penalty killing defensive pair featuring him that gave up a goal.
Simon Nemec has played more on the penalty kill in his career, logging 69:22 over the last two seasons in that role. With only five power play goals allowed in that time, Nemec should be played more on the penalty kill to see if this can become an aspect of his game. However, Nemec has struggled with defending the net, and Brad Shaw will have to work with him in that area of the game.
Last Year’s Units vs. My Proposed 2025-26 Units
In the 2024-25 season, these were some of the most common penalty killing groups:
- Hischier-Mercer-Dillon-Kovacevic: 77:45 TOI, 1 GF-15 GA — 11.58 GA/60, 10.32 xGA/60
- (Center)-Bratt-Siegenthaler-Pesce: 33:43 TOI, 0 GF-3 GA — 1.39 xGF/60, 5.34 GA/60, 6.08 xGA/60
- Hischier-Mercer-(Defense)-Pesce: 31:34 TOI, 0 GF-6 GA — 11.4 GA/60, 9.98 xGA/60
- (Center)-Bratt-Dillon-Pesce: 16:06 TOI, 3 GF-0 GA — 2.09 xGF/60, 8.72 xGA/60
The real key to last season’s penalty killing success seems to revolve strongly around Brett Pesce, who elevated pretty much everyone who was on the ice with him. Only Nico Hischier did not work well with him, but I do not think this is a chemistry thing. Rather, it may simple be misuse to have Hischier on the penalty kill, beyond taking faceoffs. This is how I would try to line everyone up in the 2025-26 season:
- Glass-Brown-Siegenthaler-Pesce
- Hughes-Bratt-Hughes-Hamilton
- Hischier*-Palat-Dillon-Nemec
- Mercer-Noesen
Perhaps some of Hischier’s previous season is bad luck. How many times did he win a big faceoff draw, only for Dawson Mercer to fail at lifting the puck out of the defensive zone? This could factor into how Hischier is used this season. He is very effective in the dot, so the team could use him for defensive zone draws in place of Jack Hughes. Then, the moment the Devils get the puck out of the defensive zone, Hischier should get off the ice. I did not specifically keep track of this in 2024-25, but I felt like Hischier stayed on for extra-long shifts far too often on the penalty kill, which may have contributed to his poor results. This was also true of Dillon and Kovacevic, even if the Devils successfully cleared the puck down the ice. Hischier is exceptional at remaining a top offensive threat even 45-60 seconds deep into a shift (yes, there is data on this), but this only does so much good in a situation that is focused on defensive zone play.
You might be concerned about that second defensive pair. However, I do not think it would be a good idea to pair Hughes with Nemec. The weak points in their shot effects overlap (i.e. a heavy increase in shots below the left-wing faceoff dot), and they are both still very young. I also do not think Dillon has shown that he should be a top penalty killer and would be better off focusing on making an impact at five-on-five. Comparatively, Hughes and Hamilton have different weaknesses. Hughes may be the fastest defenseman in the NHL and has a great defensive stick, but he can drift a bit and can toughen up around the net. By comparison, Hamilton is giant and effectively takes away a lot of space as a result. And if he is on the penalty kill, I doubt even NHLers would be brave enough to step in front of his clearing attempts.
The biggest thing is, in my opinion, keeping clear top six guys out of the top penalty killing role. Cody Glass has excellent five-on-five defensive metrics and did well in career-high penalty killing minutes in 2024-25. Connor Brown is a lifelong penalty killer, and he can take a lot of tough minutes away from more productive players by continuing that role with the Devils. Ultimately, the penalty kill is a situation where teams expect to take some goals against. It is not only best to limit those goals against, but also to make sure that top scorers are not tiring themselves out in minutes where they are least likely to light the lamp.
Think in terms of momentum. Say Nico Hischier has been on the ice for a penalty kill shift, winning a draw, the clearance of which is botched by one of the other three players on the ice. The Devils are hemmed in, giving up shot after shot. They survive at first, but Hischier stays on the ice, with Bratt coming on for Dawson Mercer as the second penalty killing unit, only for the opponent to score a goal with under 30 seconds left to kill. For reference, Hischier and Bratt saw six goals against in just 17 penalty killing minutes played together. After that happens, who can the Devils send out? With Hischier out of gas, and Bratt coming off a penalty kill goal against, maybe the team turns to the third or fourth line. Last season, the bottom six would probably fail spectacularly at trying to shift momentum after that goal against, spiraling the Devils into several minutes of uncompetitive hockey.
The goal is to be competitive by maximizing the skills that players are best at. If Hischier exchanges two or three penalty killing shifts per game for even strength time, you might see his scoring tick up towards 80 or 90 points. He might even be that momentum stopper after the Devils do give up power play goals. Conversely, Cody Glass would best earn his money by allowing Hischier to take those shifts off. Getting scorers like Jack Hughes and Jesper Bratt on for those secondary killing shifts should give them plenty of time to hunt for shorthanded scoring chances, as Jack grows more into an all-situations player.
In the absence of Kovacevic, finding extra minutes might be tough, but Dougie Hamilton has had success in that role before. I also think that, with the right coaching, Luke Hughes can use his quickness and reach to become an effective penalty killer, while Nemec fixing his netfront defense would turn him into a bona-fide first or second pairing right-handed defenseman. Before all that happens though, I am just very thankful for Brett Pesce.
Your Thoughts
What do you think of the Devils’ penalty kill? Do you think Brad Shaw will be able to bring them to new heights? What do you think about the deployment suggestions I made? Do you agree with the pitfalls and strengths I highlighted? Leave your thoughts in the comments below, and thanks for reading.
Many thanks to HockeyViz for their shot map visualizations, to Natural Stat Trick for their data tools, and to the Devils’ media team for posting so many goal videos on YouTube in the 2024-25 season.