All Three Zones Overview: Zone Exits

We are now at the final portion of the All Three Zones project, which covers zone exits. This is probably the most complicated category to track because even though zone exits are regarded as an important part of the game, everyone’s had a different way of tracking them. They also haven’t been studied extensively, so there are some questions regarding their reliability as a stat. That said, they’re still regarded as an important part of the game.

If you ask most hockey minds what the most important skill is for a defenseman, one of the things at the top of their list is the ability to exit the zone cleanly and start breakouts. There’s no arguing that. I gave a few examples using the zone entry and the passing data of how some of Minnesota’s offense started with making plays out of their own zone. Conversely, I also showed how Arizona’s struggles to exit the zone cleanly contributed to them not generating much offense. Those are just examples, though. The impact that zone exits have on stats that matter like shot attempts & goals is a bit of an unknown at this point.

There’s no debate that zone exits are important and being able to do it with possession of the puck is a great skill to have, but the method of tracking them hasn’t been agreed on and it’s led to some varied results. In my 2013-14 project, my method was to track every single defensive zone puck touch, track every exit and how it was done (whether by carry, pass or clearing it off the boards or the glass). Turnovers and exits that led to icing were also tracked in this study. This created some noise when it came to analyzing the data.

My usual way of looking at it was seeing how often a player exited the zone relative to how often he touched the puck (Exit%) and how often he did it with possession (Possession Exit%). This logged all touches in the defensive zone as equal and probably wasn’t the best way to go about it. I also never tested this data but some other bloggers looked at my 2013-14 dataset & didn’t find too strong of a correlation with Corsi or meaningful possession based on my work.

This doesn’t mean we should throw out zone exits, because it’s an important part of the game, but how they are tracked probably needed some cleaning up. Thus, I’ve made some changes for this coming season. You’ve seen most of them in my playoff tracking and I’m making a couple more additions to make it better.

First off, instead of tracking all puck touches in the defensive zone, I’m only going to be tracking ones that are attempts to exit the zone. This will hopefully cut down on some of the noise and give us a better idea of who has the biggest workload on their team and who the most efficient puck movers are. As for what defines an exit, that is staying the same. Anytime the exiting the team gets the puck over their blue line, it is logged as a successful exit. The method of tracking certain exits has changed a little, though.

I’m still tracking whether the puck was carried out of the zone or passed to a teammate (these are logged under Exits with Possession) but exits without possession are defined a little more specifically. I’m now tracking pucks dumped out of the zone & recovered by a teammate (Dump-outs) and pucks that are simply cleared out of the zone (clear outs). I’m also tracking failed zone exits, which were not logged in the 2013-14 dataset, and turnovers are lumped into this group, as well.  Finally, I’m also tracking the area of the ice the exit originated from and if the exit led to a successful zone entry. The latter is to show which exits are leading to potential offense and if there’s any repeatable talent there.

In the end, we’re left with these categories which I’m going to reference often:

  • Attempts: How many times a player attempted to exit the defensive zone.
  • Exits/Exit%: How often the player successfully exited the zone, regardless of if it was with or without possession of the puck.
  • Possession Exits/Possession%: How many times a player carried the puck out or passed it to a teammate to exit the zone.
  • Carries: How many times a player carried the puck out of the defensive zone.
  • Passes: How often a player passed it to a teammate to exit the zone.
  • Dump: How many times a player dumped the puck out of the zone AND had it recovered by a teammate. This is an exit without possession.
  • Clears: How many times a player cleared the puck out of the zone with no recovery.
  • Failed Exits: How many times a player failed to exit the zone on an attempt. Pretty self-explanatory.
  • Icings: Also self-explanatory
  • Pressure: If there is an opposing player within a stride of the player exiting the zone, it is marked as a pressured exit.
  • Area: The area of the ice where the exit originated.
  • Pass Targets: If the puck was passed out of the zone, this is the player who the pass was being intended for.

Looking at Possession Exit% is my preferred method of seeing how effective a player (particularly a defenseman) is at zone exits. Being able to exit the zone with possession is the optimal scenario for most teams and you ideally want players who can do this efficiently (i.e. exit the zone cleanly on most of their puck touches) and then it into offense the other way. Let’s go to our game example to see what we can dig out of that.

Arizona Coyotes Zone Exits


I’m only looking at defensemen because the sample size for the forwards is pretty small. Arizona’s blue line had a bit of a tough time exiting the zone cleanly, as I pointed out in the zone entry article. Connor Murphy was the only player who exited the zone cleanly on more than 50% of his attempts and he split his role pretty evenly with Nicklas Grossmann, who was easily their least effective player here. A little surprising to see Ekman-Larsson rank so low here, but it’s only one game. Another thing you’ll notice is Zbynek Michalek having more exit attempts relative to his ice time than any other blue liner. This is mostly because he did most of the work on his pairing with Klas Dahlbeck. Michalek’s not exactly known as a puck-mover but he did an okay job here relative to his workload.

This will look better once we have more games tracked and have a better idea of what the league average success rate is for a defensemen, but this is a good starting point. Just remember that we’re still in the beginning stages here and take the numbers with a grain of salt. Ekman-Larsson’s numbers also got me curious. He’s regarded as a very good passer, but had only three successful passes on 13 attempts (we’re subtracting his one carry-out). Was it bad breakout passes or the forwards mishandling them? Since we’re tracking pass targets, we might be able to find out the answer.

Here’s who Ekman-Larsson’s passes were going in this game:


Through this we can tell a few things. First, he wasn’t leading any breakouts when Arizona’s first line was on the ice (Vermette’s line) and he was passing the puck to grinders for the most part. We can also look at where his passes were going through looking at pass area data.


Almost all of his breakout passes were in the left lane, so he was mainly passing the puck to left wingers, which limits his options. After digging into the video, it looks like there’s some blame to share on both sides, but some of it was forced by the situation. Yes, Ekman-Larsson’s passes weren’t going to Arizona’s most skilled players but the passes themselves weren’t exactly easy to handle. Minnesota also pressured the defenseman into making quick plays, which forced the issue.

This is Ekman-Larsson making a quick pass in the middle of the ice after breaking up an odd-man rush for Minnesota. He quickly sees the forward in the neutral zone (Downie) and this is probably a breakaway if the pass is fielded cleanly. It ends up going through Downie’s skates and the play results in icing. Ekman-Larsson probably could have waited another split second to make the play so it’s easier for Downie to handle but that’s easy to say in hindsight. It’s a fast-moving game and he probably thought he could have made the pass.

Here’s what a more controlled breakout looked like:

Ekman-Larsson had the puck on his stick for a few seconds here, and he goes with a pass up the wall to the right winger. Minnesota’s forecheck looks aggressive, but neither forward was putting a lot of pressure on him so he had time to make a play happen. Unfortunately, his pass went off Chipchura’s stick and right to the Wild defenseman, so this ended up being a dead end. The Coyotes probably should have gotten possession here even if the pass wasn’t completed cleanly, but the Chipchura had some problems with fielding this.

The Wild picked up on this later in the period and continued to pressure Arizona’s defensemen on breakouts, especially after faceoff losses.

Arizona just won a faceoff here, and you can see the Minnesota forward go right for Ekman-Larsson and force him to make a play. He doesn’t have much choice but to go up the wall here and the Wild quickly get possession back because Chipchura couldn’t handle the pass. Pressure forces bad breakouts and Coyotes fell victim to that multiple times this game, Ekman-Larsson especially.

So, there are some external factors to consider with zone exits. I’m hoping to account for all of them, though as there’s a lot I’m tracking here and hopefully we can get a better idea of who the best & worst puck-movers are through this study.


Again, this is something new I’m working on, so remember to take that into consideration when going over the data. Think of it as a “bonus” with all of the other stuff I’m tracking. Still, zone exits are a pretty important part of the game and I’m hoping that tracking them through this method leads to some meaningful information that we can use to evaluate players.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s