Caps PK


If you ask the NHL, the Washington Capitals penalty kill was the 16th best in the land this season. They allowed 51 goals and escaped 82% of their shorthanded sessions unscathed, which doesn’t sound so bad. But actually the Caps PK was really, really, really, really bad. They were one of the worst ever, as far as I can tell.

I’ll start with the good: Capitals goaltenders had a .906 save percentage while playing 4v5, which is way above average and good for 3rd best in the league. Most of the credit for that should go to departed goalie Jaroslav Halak.

Goalie Shots Faced 4v5 Sv%
Grubauer 70 .957
Halak 234 .910
Holtby 237 .865

Update: Halak’s numbers include his time with St Louis.

Though Braden Holtby was a bit below average, Washington’s goaltenders did a respectable job masking how fantastically awful the Caps PK unit was at suppressing shots. Cause they were like historically awful at it.

For every sixty minutes the Caps spent on the penalty kill, they allowed 89.5 unblocked shot attempts– far and away the highest number in the league. Actually, any year. As far back as I can see, I can’t find any team with a higher shot rate while a man down.

The chart at top, the creepy one with Adam’s face at right, is based off ExtraSkater’s fenwick data since the 2011-12 season. The Caps saw a higher rate of unblocked opponent shots than every other team. They’re ranked 90th out of 90, and they were head and shoulders above number 89.

Using Behind the Net, here’s a frequency distribution of 150 teams since 2008 using just opponent shots on goals per 60 minutes.

caps pk3

Three more seasons of data, this time excluding missed shots, and they’re still the worst.

Now, just because I’m enjoying this, here’s a scatter plot of opponent corsi events — including misses and blocked shots– per 60 minutes since 2007. These data are from Hockey Analysis.

cap pk2

Yeaaaaaaaaaaah. Playing a man-down, the 2013-14 Caps were a special snowflake of suck. You really gotta feel for the goalies.

I’m not 100% sure, but I suspect the biggest difference between the not-god-awful Caps PK of the pre-Oates era and the make-it-stop-oh-god-my-eyes-please-make-it-stop Oates PK is aggression. Before last season, it was more common for the Caps to attack the puck when it moved from the point to the half boards or vice versa. In 2013-14, Calle Johansson (who is technically still employed) instructed the boys to stay back, opting to clog up passing lanes instead of challenging the puck.


“We pretty much give them all the time in the world,” Karl Alzner told your boy Chuck Gormley in November, “and just try not to give them any lanes.”

That didn’t work at all. We knew it then. I find it comforting how reliably these “sit back” strategies fail.

Adam Oates, known for his attention to detail as well as his penchant for losing, was particular about keeping his even-strength forward pairings together when on the PK. His justification was that they would “know where each other is going to be on the ice at all times,” which you’d think would be pretty easy since they hardly ever moved.

Look how Brooks Laich and Troy Brouwer prioritize occupying the slot over forcing a turnover along the boards. When Claude Giroux gets his shot off, he has a bunch of space and just two Caps penalty killers in any position to challenge him.

On this next one, Karl Alzner does something similar, abruptly halting a chance to take the puck along the boards so that he can stay near the slot, which doesn’t even help since Wisniewski puts a dart past him right after.

I’m sure there’s a lot more going on here– perhaps related to how the team responds after faceoff losses– but I’ll defer to the real stat geeks and X’s and O’s guys to diagnose it. Considering how extraordinarily bad this team was at suppressing opponent power play shots, it seems like this deserves closer examination.

But if we never find out exactly why the Caps PK was so bad, and they just go back to some un-awful shorthanded system next season, I’ll probably be okay with that too.

Tagged with:
  • those are perfectly cromulent stats.

  • Brouwer Rangers

    You’re forgetting that this strategy means the only shots opponents get are from the perimeter and there’s no reason of any kind to fear perimeter shots. #duh

  • fingerman

    There’s no way Halak faced 234 shots at 4 vs. 5 with the Caps–he only played, what, 12 games with the team? If that was the case, the Caps penalty kill was truly the worst ever assembled by any hockey team in history. Would presume he faced similar amounts of shorthanded shots as Grubauer, which means Grubi should get as much credit–even more–for the great 4 vs. 5 shot percentage.

  • He saw only 383 shots total with WSH. Lemme double check what I got from Hockey Analysis. They might not have broken it up by team.

  • standarsh


  • I added a clarification.

    Grubi was a stud for sure, but with only 70 shots, he wasn’t moving the needle much.

  • shut up

  • fingerman

    Sure–but Grubi played 17 games. He must have moved the needle as much as Halak did in his 12 games.

  • fingerman

    The stats, though, are amazing, whoever was in goal.

  • Brouwer Rangers

    But seriously – is shot quality (if I’m using the term correctly) to blame for such a horrible PK in terms of shot attempts allowed actually being just a mediocre PK in terms of goals allowed and kill percentage?

  • JH

    I seem to recall, but cannot find it anywhere at the moment, that Oates didn’t want his defenders blocking shots (PK or otherwise) but wanted to essentially err on the side of letting the goalie see the shot coming? Of course that means that the shooter has a great view of the net, too. Maybe it just seeped into my unconscious after watching horrible defense all year?

  • Graham Dumas

    No, I remember that, too. Cannot recall from where, though.

  • tpr04

    Good stuff on the PK. The other thing that you have to consider is that the Caps were a bottom 10 team in terms of taking minor penalties. You add that to the equation and it makes a bad situation worse.

  • alchemistmuffin

    “But if we never find out exactly why the Caps PK was so bad, and they just go back to some un-awful shorthanded system next season, I’ll probably be okay with that too.”

    No I am NOT okay with that. We need to find out what caused out PK to go downhill this year, in order to make changes. Because I can tell you, based from what I’ve seen, the other teams in the Met division will take full advantage of that next season when they retool their PP systems against the Caps.

  • VeggieTart

    From what I understand, Oates didn’t have that “killer” instinct. Even at even strength, he was pretty much a sit-back-and-hope-they-make-a-mistake type guy. His style was more suffocating than Dale Hunter’s, and that’s saying a lot.

  • jhershb

    Excellent review. The worst part of watching the PK this year from up in section 404 was that the Caps almost never seemed to contest the puck in the neutral zone, and instead simply backed up and let the opponent carry the puck into the offensive in a relaxed fashion. Other teams were not so hospitable, and often broke up Caps’ efforts to carry the puck into the zone, forcing them to dump and chase. Let’s up the new HC changes this.

  • Barrett

    His style isn’t more suffocating than Dale Hunter’s style….he just suffocated the wrong team. Hunter employed his best defensive lines to attack the opposing teams better offensive lines. Oates just told all his lines to not play defense.

  • vahockeymom

    I wish I could like “suffocated the wrong team” 10,000 times!

  • VeggieTart

    Well, I meant suffocating of the Caps, of course. Yes, Oates suffocated the Caps, but Hunter’s style was a flawed too.

  • Owen Johnson

    It would be cool to see a heat chart of shots the Caps give up VS a heat chart of what, say, Boston gives up. Although, lower quality or not, the fact that we give up that many shots more than anyone else has to be worrying.
    Also, reviewing those PK videos, a lot of the Caps are doing literally nothing.

  • Diller M

    Ok guys were gonna let them come in, set up and shoot as much as they want, now I don’t want you to try and block the shots, because that might screen the goalie a little, so be sure to get out of the way when they shoot. You’re job is to stand in one place and move a little left and a little right, if they try to pass it and you can reach it with your stick try and get it, but be sure it’s not a shot, because I don’t want any weird deflections. If you don’t get the puck just stay put until they try to pass it by you again. Oh and be sure not to chase after loose pucks because that will make it harder for them to get their unblocked shots off.

  • Lawrence

    You can’t discourage shot blocking and expect to be a good team on the penalty kill.

  • John Knab

    The analysis here is flawed. The only key stat is that the PK was 16th in the NHL, meaning that only 15 teams were better. Oates’s style was no great innovation – it’s been around for at least 35 years, when I saw it in college. The idea is to keep shots to the outside – to the low percentage shots. The stats you use show the style had success: because the other team took many low percentage shots, the goalies’ save percentage was higher. Connected to that, there’s less need to block low percentage shots. So far, everything makes sense. You and others may complain you don’t like the passive style, but you can’t use stats to paint a partial picture.

  • Owen Johnson

    I think you are right to an extent, but when a team gives A TON more shots than the next place team (which is the freaking OILERS), it’s pretty troubling.

  • Diller M

    So the question then is, would you be OK relying on this same PK next year?

  • Sarah

    Halak’s time with the Blues is included in those numbers?

  • Bilal

    what cauused our PK to go downhill this year?
    easy. The coaching staff

  • Larry

    Please people, please don’t drink all this Kool Aide! There is a reason why REAL hockey people always say “stats are important, but don’t tell the whole story”. The rest of the story is completed with a strong in-depth knowledge of the game. It is not completed with more stats! Corsi, Fenwick and all the rest fall into this category. One cannot compensate for a lack of true hockey knowledge by spending more time on stats. Do you think Montreal was thrilled to be going into the second OT last night being outshot 50-29? SOG’s are a pretty important stat, and have been for a long time, but as I said, don’t tell the whole story! Subban has an absolute cannon, and it only took 1 shot to win that game at that point.
    Caps PK (and season really) can be attributed first and foremost to a very young, thin defence that had experienced forwards trying to do more as they tried to compensate. Being so tight to the salary cap meant defence call ups had to be selected by income and not ability, absolutely no bearing on coaches or system. (Some of the players had been recent healthy scratches with the Bears prior to the call, not exactly the “next best thing”)
    When trying to solve a problem don’t get tricked into addressing the symptom (coaches and system), rather consider the root cause.
    Oates is gone, you have stated your opinion on him, but to say stupid things like “the worst coach ever” and “may never work again in hockey” devalue your blog as it’s clear the author of the top ranked PP in the league two years running quite simply can’t be that dumb. And before you bandwagonners get typing, don’t fool yourself for 1 minute thinking that was all Forsythe, that system was never anywhere he was previous, I doubt he would wait until Oates arrived to reveal it.
    Thanks for taking he time to read this.

  • Fred Merc

    Yep, we should fire him!
    Dead horse sayeth: “Please stop beating me!”

  • kyle boyd

    i’m okay not pressuring the halfboards (depending on the opposing PPs structure). i mean think of how often we see backy pressured along the halfboards, slip it to green, who exploits the aggressive play and finds a lane to ovy when they scramble to get back into position. i honestly think being soft on at least one of the 2 players playing the half boards is the best way to counter the 1-3-1.

    the problem here (beyond how easy we make it to enter the zone clean, which i honestly believe warrants its own discussion)is that we’re not doing the ‘passive approach’ when you take this more passive approach its not that your letting them take shots from the perimeter, its that you’re letting someone control the puck. a guy from the half boards, especially if he’s angled low, isn’t in a favourable shooting. our problem is that we’re not taking away his lanes to the top of the umbrella, and then when it gets there we’re not being aggressive at the top of the point. why do both laich and brouwer get so low when it goes behind the net? (not criticizing either i’m sure this was a coaching decision). Leaving any of the guys outside with clear shooting lanes when they’re at range is okay if thats the cost to take away passing lanes and we’re in position to collapse on them should they try and walk it in, but we often times did neither.

    the ‘passive approach’ is supposed to just be controlled aggression. what we were doing in that philly clip is like nothing i’ve seen in the 20 years i’ve played.

  • Shot quality is kind of a red herring when you’re talking about only 500 shots or so. At that sample size, random luck is a much bigger factor than a team’s “talent” at limiting shot quality.

  • Yeah, I don’t wanna overemphasize the TACTICAL decisions that led to the insane volume of shots. Maybe those tactics are valid or maybe they could executed better. The important thing, to me, is that they didn’t work this time around.

    The subtext here (and in a lot of things I’ve written and will write this summer) is that the team will be better next year.

  • If you can separate the apples from the oranges, I’d be delighted. I don’t have that data readily available!

  • I don’t think I’ve ever failed to give Adam Oates credit for running the best PP in the league, but power play is still like 15% of the game tops. So maybe he’s just 85% awful? I’m okay with that if you are.

    Thanks for leaving a comment. It definitely undevalued the blog.

  • kyle boyd

    ye i actually don’t feel to bad about the future. to me the most frustrating part of this season was how glaringly obvious and seemingly simple to fix our issues were. considering that however, next year we should be able to easily right the ship with the right guy(s) at the helm. most of my friends point to the caps D as a real weakness but i think within the 14 we played this season are 6 who can do a great job (and hopefully we can use some cap space to sign another)

  • RESmith

    MY GOD! The concentration of all those shots just on top of the crease is downright scary. It is as if the Caps just stood there and served more as an obstacle course than a PK unit.

  • Owen Johnson
  • Owen Johnson

    Wait. I just realized that chart measures when WE are on the PP. Here’s Holtby’s side by side with Steve Mason (who according to that graph had the “best PK” this year). It’s not nearly as damning. In fact, the other teams shots ARE on average farther away

  • Eric Schulz

    I think that, sure, we have 6 that can do a great job, but not 6 that can do a great job while complementing one another. I’d say we have 4 definitely, and a few guys who could serve as solid 7ths… it wouldn’t be surprising if at least one of them improved enough in the offseason that we had 5, but we definitely need to add one… shouldn’t cost too much either, as generally it’s the point producers who cost more, and we just need a steady guy.

  • The Fresh Prince of Nowhere

    Was too funny not to share

  • Sarah

    Sorry, not meant to be a complaint. I love fruit salad as much as the next person.

  • Sarah

    It’s legal in Colorado now…

  • Pat Magee

    Let’s just please stop looking back at how bad we were, and just look forward. Let’s see the possible coach and GM candidates!

  • Josh Carey

    Do you think that Oates could draw up and equally good powerplay without Ovechkin? Sure there are some teams that have a guy that can distribute like Backstrom, but many of the Caps PPGs (I’d argue more than half) came because of Ovechkin. Either his incredibly accurate slapper, which I don’t think anyone in the league can hit the net as much as he can from “the Ovi spot,” or because a team left a guy on Ovi creating pretty much a 4on3 on the right half of the ice.

    Oates did a good job of making a good PP, but could he create a top 5 PP without Ovechkin? He certainly couldn’t in New Jersey, and that was with Kovalchuk.

  • johnnymorte

    I have reached the conclusion that the only coach who could take this team to the promised land is Mike Keenan. If they don’t try to get this man than it is a fool’s gambit. He is fourth all time in playoff wins and he just resurrected himself in Russia. Most importantly, he is a direct pupil of Scotty Bowman working under him in Rochester.

  • Mike Logan

    Keenan is a high wire act – good coach, who thinks he’s a GM and insists on control over player personnel; his teams invariably are gutted of talent once he’s worn out his welcome. Not to say the caps couldn’t use a roster makeover – I just don’t think Keenan is the guy you want doing it.

  • Mike Logan

    NJs pp was horrid before Oates got there.

  • Josh Carey

    If Oates is as good as everyone says on the PP (which I admit, he’s pretty good), he should have been able to do better than 14th in 2011-2012 with Kovalchuk, Elias, Zidlicky etc. The fact that those guys were horrid before Oates just tells you that the person before him was really bad. Any PP with Kovy/Elias should be able to run along at 20%+

  • johnnymorte

    I respect your opinion regarding Keenan’s methodology, however I strongly disagree. The reason why Keenan was so successful in Russia is that many Russians respect old school thinking. I believe he can get the best out of Ovi and a lot of other guys that have gotten way too comfortable here. Keenan loves an uptempo physical game and skilled guys flourish under him more often than not. He’s going to upset a lot of players, but this team needs a gut check and Keenan will ride them until they break. You look at Scotty Bowman who was a master at getting underneath a player’s skin, Keenan was under that tutelage. We’ve already seen that the player’s coach mentality does not work with this group. Rather than blow up the team, get someone in here with a different.perspective that has done it before and see what the result will be.

  • Jonah

    So not only are they letting up more shots? But there’s more shots from the INSIDE slot?
    I thought those were the only shots that mattered and our whole point was to stop those?

  • Jack Soule

    Here is the problem–the Caps actually look like they were taught this system well. They did what they were supposed to do. This system is designed specifically to block one-timer passes from point-to-circle, circle-to-circle, and cross-crease. With the penalty-kill, it really comes down to picking your poison–each system has its benefits and its flaws. I can absolutely understand why and how Oates came up with this system: he worked very closely with two of the best Russian one-timers over the past three years. Kovalchuk and Ovechkin are two of the best one-time shooters in the game–there are others who are good but lets just all agree that they both have sick shots–and after seeing your own team score primarily with one-timers (as opposed to deflections and second chances) in a 1-3-1 setup over and over and over, I can understand the desire to block one-time passes.

    With that said, watch those two videos again and you’ll see that each player is doing exactly what he should and–if you want me to draw up X’s and O’s to confirm, I can–is exactly WHERE he should be. In the Philadelphia video, Brouwer is standing in the passing lane of point-to-left-circle (Timonen to Giroux) since that is where the puck is and the easiest pass for the opposition PP at the moment. At the same moment, Laich is covering the pass to the other circle (Giroux to Voracek). Then the two defensemen are left to play two-on-two down low. Carlson is ready to block the shot and also take away the pass down to Hall to the left of the goal. Alzner is then left to deal with crease duties, or to get into the shooting lane should the puck come over to Voracek.

    On to the video against Columbus: Alzner did as he was coached to do by letting up off the puck. He returns to the slot to block a pass into the slot or cross-crease. Basically, he is in position to block a pass that would be coming from either Mojo down low or Nicky on the half-boards over to Ovi slipping down from the top of the circle to between the bottom of the circle and the dot. Does that make sense? Basically, this penalty-kill was designed to stop Backstrom, Green/Carlson, and Ovi hookups. Therein lies a small part of the problem and the first part of the problem.

    To sum up–the Caps penalty-kill was designed to block the kind of one-timer passes that would happen on the Caps power-play. The problem with this is that it is a preference for an attempted shutting down of the top portion of a 1-3-1 or umbrella formation. Sure, when you have Ovechkin or Kovalchuk on your team, you want to take away that pass. But against every team? No. Hell no. Most teams, unlike the Caps, recognize that a good power play outworks the penalty kill. The Caps power-play is built around playing pretty hockey, and although it makes my hockey 101 instincts cringe, it works when you have Backstrom and Ovi. However, most teams–especially the Flyers–look for deflections and rebounds.

    In theory this may be a good system. In real life, on ice? Not so much. On real ice and in real hockey, pucks go wide of the net and rebound off the board, pucks are shot wide and go off sticks legs and skates, pucks come off a goalie’s far pad and into the slot, this can happen, that can happen, and the other can happen. We all love watching Ovi’s one-timer goals but most great playoff team thrive on creating controlled chaos–get bodies in front of the net and just chuck the puck anywhere near the goal and good things can and will happen when you outnumber the opposition in that juicy area called the slot.

    So, problem number one: this is a system that is theoretically based on stopping perfect hockey from happening. Well, folks, hockey involves a lot of crazy **** happening. In other words, this system tries to take away the umbrella part of the 1-3-1 when in actuality, despite both videos above being examples of a goal that should have been prevented in this system hitting the twine, a lot of the best and truest hockey most frequently occurs down low, at the handle of the umbrella, if you will.

    Problem 1(b): the top part of Oates’ PK formation might work as long as the puck is pressured down low or when the puck is loose. But the Caps showed us that Oates more-or-less told his players to NEVER pursue the puck.

    Problem number two: this system is not actually completely idiotic as I may have implied directly above, what is though is a coach discouraging shot-blocking. This setup would work decently well if it was played with more heart. But all Oates systems are fairly robotic and heartless. You don’t need to play Hunter’s controlled aggression style, or Boston or Philly’s all-out aggression style to frustrate the other team and create turnovers. Building a system to counter the most commonly utilized 1-3-1 setup COULD work. But the team would have to play with the heart that we saw from Columbus last week, not the Caps team-defining lackluster attitude. IF the team was willing to sell-out to block shots as well as passes–even if that means risking being briefly out position–it COULD be effective at keeping the opposition to the perimeter and an NHL caliber goalie is going to stop almost every shot that he sees that is a straight-shot and not a one-timer. SO Oates system was not terrible, the fact that he told the players to intentionally not block shots WAS.

    Problem 2(b) is that along with being taught to not block shots, I absolutely hate the current trend in coaching of encouraging defensemen to stand in front of enemy players in the crease, not behind them like in the good old days. Sure, let them get in front to block the shot IF you are playing a shot-blocking system. But if you are playing a system that is built around LETTING the goaltender see perimeter shots with minimal risk of deflections, then the two defensemen playing down low need to clear the crease. So, defensemen’s positioning is okay but not my favorite IF the shot is attempted to be blocked. If the intention is not to block, then play good old-fashioned boxing-out defense.

    Problem three: as much as Oates seemed to like tracking scoring chances (since he believed that was the only stat that mattered and left shot volume and possession by the wayside), this system allowed for soft spots within the scoring chance zone (primarily the high slot). However, I believe these soft spots were due more to the lack of Caps player movement then the actual positioning and underlying philosophy. The biggest weakness I saw was a pass from the goal-line-extended to the middle- to high-slot. This would occur when the PKers would back off the puck in the corner: as the PP puck carrier would swing up from the corner toward the half-board, the defensemen would collapse back toward the front of the net while the forwards moved up toward their positions to get in passing lanes at the top of the PP formation. This simultaneous movement left all four Caps player either stationary or moving AWAY from that particular soft spot in the middle- to high-slot area.

    Picture this–the puck is in the corner and the Caps are mostly covering the low slot, then the puck carrier swings up toward the half-boards and the strong side defenseman backs off toward the net and away from the corner, the strong side forward moves away from the net and toward his position to get in the circle-to-point passing lane, the weak side forward moves from the low slot away from the net and toward his assignment in the cir

    Problem four: the overall lack of effort and intensity in Oates’ systems. Man oh man, Oates was a great player but 99% of people have 200% more emotion and 200% less brain-gears whirring all at once. On paper, there are things that might work with this system. This system honestly could work, fairly well I suspect, given the following:

    One: you have a good and consistent (doesn’t need to be great, just consistent) goalie who WILL stop the first shot.

    Two: you play a bit more of a pressure system anytime the puck is below the circles.

    Three: IF you are going to play a system that allows the goalie to see more shots, then commit the defensemen to clearing the crease since they are already playing two-on-two down low and allow the forwards to cut off one-time passes in order to force long range shots that the goalie can easily stop. Long range shots aren’t going to get past an NHL goaltender if there is no screen or deflection.

    Sorry this is long and thanks for reading. This is my honest assessment of all the grey areas behind the PK–why Oates used it, what it was intended to do, and how it could have been improved from paper to ice.