An Introduction to my Zone Entry Project

As the game of hockey evolves and so too do our methods of analyzing it, the frontiers on which we plant our flags becoming increasingly foreign.

These incremental steps have been made by measuring the way in which players direct the puck on net. A game once analyzed almost entirely on the virtue of goal differential and the players that drive it, the paradigm has shifted further than ever from the goal line as we attempt to observe the signal for the noise.

This linear logic isn’t confined to the offensive zone, either. If the key to refining the way in which we observe shooting talent extends to how the puck is directed on net, then naturally the next step forward is finding out which players are driving their team into the position to make these plays.

One can’t threaten the opposition from the neutral zone, so finding out which players drive play into the offensive zone is of immense importance. Taken a step further, it’s worth the time to find out how they are pushing play on their opponents.

To that particular end, some of the more innovative minds in the field of hockey analytics saw fit to track zone entries and exits. The concept is simple: figure out which players are bringing the puck into the offensive zone and how. This exercise proved that by carrying the puck into the offensive zone, any given team can expect to nearly double their expected shot volume as opposed to a dump and chase play.

As with anything, the probability in this instance certainly doesn’t qualify as destiny, and there have been more than a few teams to buck the trend – Los Angeles Kings and St. Louis Blues in particular. All the same, their sizeable data set confirms what most of us already knew intuitively. How can one expect to increase their share of the meaningful possession pie if they throw the puck away en route to the offensive zone? A tired and fruitless exercise, if you ask me.

[A paper published by Eric Tulsky for Sloan Sports Conference on the value of tracking zone entries]

This project was taken a step further in the 2013-14 season when Corey Sznajder decided he would track every single game of that season and make the data available to the public. Given that the NHL doesn’t include this data in their play by play, that meant manually tracking every single game in Excel and synthesizing the data for our analysis and curiosities sake. Honestly, I can’t think of any more charitable an act to the hockey community, and I’m often baffled by the lack of mass attention given to this project. Similarly, not nearly enough is made of the time and resources that were dedicated, either.

To track these entries, they were divided into four separate categories: carry-ins, dump-ins, other and failed. The qualifiers make intuitive sense, but ask any two trackers for their taxonomies and you’re likely to get any number of different answers. Generally speaking, though, it naturally breaks down into how the puck was brought into the offensive zone. 

Not-so-surprisingly, these efforts landed Sznajder a cushy NHL gig – much like his predecessor, Tulsky. Which brings me to my incarnation of this project. 

This season I will be tracking every single game of the NHL season. I don’t expect I’ll be able to keep up with the season as it plays out, but I’m doing the best I can with the limited time and resources available to me. I will, however, be tracking the Canucks as they go. So far, I’m nearly four full league days through the season with any number of random games thrown in there as well. In total, there’s probably a lot closer to 5-6 days worth.

I’ll be using the same formula as the pioneers in this field, with an eerily similar template for entering and synthesizing this data. Where I vary from some of my peers is in the application of this data analysis and tracking.

The way I track entries is much more of a hybrid, touch-based system than one which arbitrarily decides the player in questions intent. For example, many trackers won’t count a dump and change play as a zone entry. I can understand the logic behind this as many teams won’t press their fortunes on the fly. For every one of those teams, though, there’s a Calgary Flames, who will be dandy on the spot, challenging the opposition. 

In keeping with this vein, I count failed dump-ins as failed entries. Seems basic enough, but generally speaking, most people will only count a failed attempt to enter the zone with control. Again, this seems somewhat arbitrary and hardly objective.

By doing this, I can increase the size of the sample without infringing on the objectivity of the data in question. If anything, it’s made more apparent. As the sample increases, so too does our ability to draw anything meaningful from the data in question.

I’m also considerably more judicious in how I differentiate the carry-ins from the dump-ins. For many who track entries, it is as simple as how a player crosses the line. Is it a controlled zone entry though if the player in question takes half a step across the line and throws the puck into the oppositions corner? I can’t think of any coach who would ascribes to that classification, so I have a hard time buying into it myself.

Of course, this might raise concerns as to the subjective elements of the data I’m drawing, and that’s entirely fair. Nearly every set of data is suspect to that element. Hell, differentiating between shots, shot attempts, blocks, etc. It’s why www.War-on-Ice.com has the option to adjust for scorekeepers bias. That’s why it’s so important to me that the sample size is increased, so as to iron out some of these wrinkles.

When it comes down to the basest levels of statistical analysis, the first task is to find out what’s happening. Only afterward can one find out what’s important. These fundamental principles are the driving factors in how I draw this information. 

Now, I’m sure you’re all curious as to how this information pertains to Canucks Army. Well, I’m getting there. In fact, tomorrow I will be posting a detailed breakdown of the Canucks zone entry data from the month of October. I didn’t want to flood you with all that information, before establishing what it means first. I hope I’ve done that with this post, but assuming I haven’t, leave any questions you might have in the comments section and I will answer them in kind.

If you’d like access to any of the information I’ve recorded, feel free to ask by email. My plan is to make this information available to anyone curious enough to ask for it. I also hope to be as transparent throughout this process as possible.

  • Ruprecht

    While the data seems to be of help, I think you are missing the player aspect.

    Sure the Sedins will be better than Kenins in gaining the neutral zone, and I think most Canucks fans want a guy like him to dump and chase.
    For guys like Bartkowski it would be interesting because he flies through the neutral zone but coughs the puck up in the offensive zone quite often.

    Then you have the style of team you are playing. What do you do when a defense first team clogs the neutral zone and forces you to dump in the puck. Forcing a carry in and creating neutral zone turnovers is the goal.

    Good luck

  • Ruprecht

    @JD

    “One can’t threaten the opposition from the neutral zone, so finding out which players drive play into the offensive zone is of immense importance.”

    You had me until this point. Maybe you want to rethink this. If a team can attack the neutral zone with speed it generates an obvious offensive advantage at the opposing blueline. To ignore this fact is leaving the very edge most teams seek out of the equation, regardless of how teams enter the offensive zone.

    You dump with speed, there’s an advantage to the forecheck. You possess with speed, then the defensive coverage is forced to adapt quicker.

    Care to defend your narrow view of the space between the bluelines here?