For the purpose of this article, penalty killing is strictly four against five. Data was either self-tracked or from Natural Stat Trick.
Last weekend I presented at the Rochester Institute of Technology Sport Analytics Conference where I used the Bruins penalty killing entry defense as my data set as an example for a project. Now I’m going to analyze the data set, and combine some video to figure out what the Bruins did well, and what they did poorly.
Before we jump in, let’s first talk about why this topic caught my eye in the first place. Prior to hosting the Detroit Red Wings at TD Garden on December 1, the Bruins allowed 110.7 shot attempts per hour on the penalty kill. That was 29th in the league, only ahead of Tampa and Ottawa. After that, the Bruins only surrendered 81.1 shot attempts per hour, which was 4th best in the league. That was a net change of 29.6 shot attempts per hour. However, this is largely due to regression to the mean. Teams that start out very bad should post better numbers later, and vice versa.
Most of the net change can be explained by the rate at which the team surrendered shots prior to December 1. However, from a day-to-day perspective, this isn’t satisfying as a fan or coach. Changes needed to be, and were made at the time, but it is important to remember that wasn’t the only driving force in this improvement.
At this point we need to ask ourselves, what causes shots? More importantly, what causes shots that the Bruins can control? My answer to that is zone entries.
Each Power-Play starts with a defensive zone faceoff for the team shorthanded. Faceoff wins can help suppress shots in this area as the team shorthanded can ice the puck, however, that has a very small impact on the entire penalty kill. A faceoff win that leads to a clear, only buys you about 15 seconds, to which you now need to defend your zone.
Of course, there is no data on penalty killing zone entry defense, so I had to track my own data set. For each entry attempt against, I recorded the initial formation the Bruins showed, whether or not the entry was controlled, and some other stuff that I won’t use here.
The differences in formation are really governed by the forwards, as the two defensemen are always going to be in the final line of defense. When the forwards stagger, like in the picture above, I would refer to it as a 1-1-2 formation. This also happened to be what the Bruins used most frequently at about 54% of the time. Furthermore, about 20% of the time there was no discrete formation for reasons such as line changes or quick regroups.
Controlled Entry Percentage by Formation
|Formation||Controlled Entry Percentage|
|Formation||Controlled Entry Percentage|
As you can see above, the miscellaneous entries (ones with no formation) concede a much higher percentage of controlled entries against. Of course, this is expected as the Bruins should perform worse when they don’t have the time to set up and counter the opponent as they would like. Personally, I see this as a contextual factor that the Bruins really can’t control for, therefore it’s not really worth the time looking into them.
If you are familiar with penalty killing systems, you are probably itching out of your own skin right now, so I must address that the initial formation and the overall concept are two different things. Looking back on it, I should’ve also tracked the concept (ex. retreating box) along with the initial formation.
Ignoring tandem pressure, an initial formation of 1-1-2 will lead to either a same-side press or a retreating box. In the same-side press, the F1 forces the puck carrier to one side of the ice, and the F2 tries to trap them.
For those familiar, this is very similar to the 1-2-2 forecheck at 5v5. The deduction of a forward on the penalty kill opens up options on the weak side of the ice for opponents. This adds a lot of stress on the weak side defenseman.
Above is a clip of two entry attempts, one without a clear error by the weak side defenseman, and one with a clear error. In the first entry attempt, the opponents are setting up for a full regroup. Both forwards start in the offensive zone (we’ll get back to that in a bit) as the opposing quarterback moves the puck to the boards quickly. This means that the “trap” by the F2 has to occur sooner, or otherwise the opponents will have too much space to work with. The Bruins want to take advantage of the opponent’s timidness as opposed to hoping they make their own mistakes.
When the F2 pressures the puck carrier around the blue line, he is forced to make a play. This is where being shorthanded is noticeable. There is an opponent where the third foerchecker would normally be. As the puck gets moved to the weak side of the ice, it is critical that the weak side defenseman stays square and maintains a good gap with the opponent. In this case, Carlo has a good gap, which doesn’t end up mattering because the pass was incomplete.
When the opponents get the puck back, they quickly regroup and we see the same formation, concept, and personnel. This time the opponents use the gap in between the F1 and F2 in order to complete a pass to center ice, which gets directed to another player on the weak side. This time the weak side defenseman does not have a good gap. He had bit on the initial pass to center ice, and is moving in the opposite direction with his back turned. At that point, the blue line is conceded. The defender must reform his gap and try to keep the opponent along the boards in order to prevent a rush shot.
Since the 1-1-2 is meant to mimic much of the 1-2-2 forecheck from 5v5 play, it makes sense that coaches should try to find ways to fill in for that lost forechecker the best they can.
In the first clip which was from October 4, the forwards both started in the offensive zone. The same holds true a few weeks later, but here the forwards are slightly farther back. This delays the “trap” to the red line. However, take a look at Joakim Nordstrom who is the F1 in this case. Rather than continuing to press to the strong side, he looks to take away options to the weak side. This almost acts like a retreating box, but the Bruins use the 1-1-2 look in order to move the opponent to the outside.
At this point you may be wondering how the Bruins dealt with the drop pass out of this formation, as the first two clips don’t have any drop passes.
One of the things I really like about the drop pass is that it forces the penalty killing team to figure out what they want to do with their F1. One of the positives about the 1-1-2 look that is shown quite well here is that it forces the quarterback to drop the puck back earlier than he would like. Ideally the quarterback would like to carry the puck all of the way to the red line. By forcing the drop pass early, the F1 gives his defensemen the time to adjust. Furthermore, the F1 forces the quarterback to one side just slightly. The drop pass becomes less effective the closer to the boards it gets.
Once the drop pass occurs, the F1 has his back turned and is away from center ice, therefore it wouldn’t make much sense to try to form into the same-side press. Just like before, the Bruins form into a retreating box and attempt to keep the play to the outside while shutting down passing lanes in the center of the ice.
The Bruins did experiment with staying in the 1-1-2 formation after the drop pass, but generally they converted to a retreating box.
When the Bruins kept the 1-1-2 formation after a drop pass, the F1 was just too flat-footed. The only thing he could really do was push the puck carrier to one side. While some teams find success keeping the same structure, it seems as though the Bruins were more comfortable adjusting into a new one.
When the Bruins didn’t show an initial formation of a 1-1-2, they were mostly set up in a 2-2. This should make sense. A 2-2 is always going to be a retreating box, but a 1-1-2 will only sometimes be a retreating box. It could be that sometimes the Bruins set up in a 2-2 accidentally due to circumstances such as a line change or the two forwards getting confused as to who is supposed to be the F1. However, these were likely coordinated after pre-scouting the opponent.
As you can see above, the 2-2 looks almost identical to what the Bruins were running. The aggressive F1 is really only necessary when there is a threat of a drop pass for reasons previously discussed. Here, the puck carrier is the last player back and has to move the puck to the boards, either right or left because the play is symmetrical. The Bruins can sit in their box and slide toward the new puck carrier after the quarterback passes the puck.
The Bruins use of the 2-2 backs up the concept that structure follows strategy. The Bruins have a set of goals that they want to accomplish and there are multiple ways to get there. However, there were a few times when the Bruins ran something completely different.
I referenced earlier that the forwards became less aggressive in the offensive zone over time. Let me just say, it’s just my hypothesis based on my own observations. I wish I knew if this was true or not. Eventually, in the last quarter of the regular season, the Bruins really did pull back the forwards as they began to use a 1-3.
The 1-3 is a very popular structure in the league. Some teams are good at it, and others aren’t. The Bruins were a team that was quite successful in this formation, only allowing controlled entries 46% of the time they had to defend their blue line.
With the F2 joining the defensemen in the back layer, it takes pressure off of them and allows them to gap up the opponent like they would at 5v5. If you aren’t familiar with how the Bruins defend their blue line at full strength, I’ve written about it in-depth here. It is worth noting that the 1-3 broke up (think turnovers) a lower percentage of entries than the 1-1-2 and 2-2. This is a small sample of data, but I believe that because the 1-3 creates pressure much later in the process, opponents are more likely to dump the puck in rather than making a mistake.
One of the keys to the 1-3 is not having the back row too far back. When the puck gets dropped at the opposing blue line, the Bruins back line is right behind the red line. Starting too far back will cause you to be flat-footed. Looking back on this play, they would’ve probably preferred to be a little farther up ice her given how slow Florida’s breakouts were.
One of the benefits of the 1-3 is that the back end doesn’t have to cover as much ice when the puck changes lanes. This makes it easier for the defensemen to stay square to the play and control their gaps. Furthermore, having three players in the back allows players to switch layers with ease. In the clip above, Nordstrom comes off of that back layer to pressure the puck carrier while Wagner covers for him in the back layer.
You can also see this switch occur earlier in the breakout as seen above. The Bruins wanted the forwards on their strong sides in the playoffs, and weren’t afraid to flip them when they expected a drop pass. The negative of doing this is that the quarterback can drop the puck back without pressure and from where he wants.
What can we learn from the 2018-19 regular season going into 2019-20? Let’s start with what the Bruins did well. When using an initial formation of a 1-1-2, they were able to force the opposing puck carrier to the outside where they would like to “trap” them. They also found a way to mimic a full strength team the best they could by assigning the other forward to the middle of the ice, creating a box. Moreover, they weren’t afraid to make small or big adjustments. They gradually tweaked their primary concept while also testing new concepts.
On the flip-side, the Bruins didn’t really have an answer for the drop pass. No matter what formation they were in, opponents controlled a significantly higher percentage of their entries. I think they might be able to find answers for that in the 1-3 where the F1 can be designated to take away the drop pass. That looks like this:
Systems, especially on special teams, are constantly changing with micro-adjustments. I am very interested to see what adjustments the coaching staff has made going into this season. Personally, I would love to see them increase their usage of the 1-3 as it was their most successful formation last season, and also has the highest potential in my eyes.