A Guide to Neutral Zone Tracking Part 1 of 2: Why Neutral Zone Statistics Matter


Frans Nielsen is incredibly good in the neutral zone.  But How important is that really?  And how can we tell that’s the case to begin with?

A while back, Hockey-Graph’s own Matt Cane wrote the following tweet:

Matt was referring to a statistic commonly found in “Neutral Zone Tracking,” which purports to measure the quality of individual and team play in the Neutral Zone.  Neutral Zone Tracking was pioneered by a bunch of guys at Broad Street Hockey (Eric Tulsky and Geoff Detweilier) back in 2011 and in the years since, a bunch of individuals have also began to do the same.  The work that’s been done on this area, on other sites as well as on this one suggests neutral zone tracking results in some extremely important data that we should be very interested in.

What is Neutral Zone Tracking:

Hockey has a variety of statistics that attempt to measure what is happening in the offensive and defensive zones – If a player allows a lot of shots against, we usually think of them as being poor in the defensive zone.  If they generate a lot of shots for, we think of them as good in the offensive zone.  We understand that these labels aren’t strictly speaking correct – a player who keeps the puck out of his dzone entirely may have better shots against than one who is simply good in the dzone – but they at least seems to work in theory.

By contrast, Hockey has no conventional statistics for measuring play in the neutral zone, which you know, takes up 1/3 of a hockey rink!  As such, the importance of the neutral zone, and who was particularly good at playing within it, was basically unknown – absent the thoughts of scouts and coaches.

Neutral Zone Tracking aimed to change that.  In NZ Tracking, a tracker tracks every time a player for each team gets the puck out of the neutral zone and into the opponents’ zone – what is known as a “Zone Entry.”  A tracker marks down the strength of play (5v5, special teams, etc.  As we’ll explain tomorrow, some trackers only track 5v5), the time of each zone entry, the player (or opponent) who makes each zone entry, and the type of zone entry.

There are 2 types of zone entries that are tracked –
1.  A controlled zone entry, where a player carries the puck in or passes it in and thus has control as he enters the offensive zone; or
2.  An uncontrolled zone entry – where a player dumps, tips, or otherwise gets the puck into the offensive zone such that he has to chase the puck down in order to regain control.

We often talk about controlled entries simply as “Carry-ins” and uncontrolled entries as “Dumps,” because those are the most common forms of those types (and because of how we track them), but it’s important to note other actions fall into these two categories.

The goals of this tracking was to find out answers to multiple questions:
1.  What is the value of a controlled zone entry?
2.  What is the value of an uncontrolled zone entry?
3.  Who are the best players at making controlled zone entries?
4.  Who are the best players at preventing opposing controlled zone entries?
5.  Who are the best and worst players in the neutral zone overall?
6.  Can we split up performance by players and teams into performance in each of the three zones?
7.  If so, how repeatable is neutral zone performance, particularly in comparison to performance in the other two zones?
8.  Wrapping it up:  How important is play in the neutral zone?

What we Found:

We found a number of answers to these questions thanks to the work of trackers for multiple teams, including Corey Sznajder who tracked every game for every team of the 2013-14 season.

First, we found that a controlled zone entry results in more than twice as many unblocked shot attempts (fenwicks) as an uncontrolled entry.  The amount for each is roughly .58 fenwicks per controlled entry and .26 fenwicks per each uncontrolled entry, but the exact amounts differ based upon who is doing the tracking.*

*We’ll talk about the reasons for this, known as tracker bias, in Part 2.

As a result, when you hear your local color commentator complain his team isn’t attempting to dump and chase more when behind in score, remember that uncontrolled entries result in significantly less shots (and of course goals) than controlled ones.  In addition, research has shown that failed attempts at controlled entries are much less harmful than conventional wisdom suggests, meaning that going for a carry-in instead of a dump-in is likely the more effective strategy in many circumstances.

Second we were able to identify that some players are particularly effective at entering the zone with control than others and that some players take bigger roles at trying to make zone entries than others (and unsurprisingly, forwards take larger roles than defensemen).  That said, it was also found, unsurprisingly, that this information on its own wasn’t completely useful – because some players may be great at making carry-ins, but may also suck defensively at preventing their opponents from doing the same (an example of this is John Tavares).  As a result, knowing who carries in a lot doesn’t necessarily tell you if a player is good in the neutral zone – we can best infer that by looking at the team’s on-ice neutral zone results while a player is on the ice.

Third Neutral Zone Data has allowed us to split up how much of a player or team’s possession #s derive from that player/team’s play in each zone on the ice.  The way this works is as follows: We look at the neutral zone results of a team and from the average results for each type of zone entry they make and they allow, we can get a score for their neutral zone performance (often called either “Neutral Zone Score” or “Neutral Zone Fenwick”).  Essentially this score measures what a team or player’s fenwick would be if that player/team had average results in the offensive and defensive zone.

Once we have this statistic, we can look at whether a player/team creates more or less shot attempts than we’d expect from their own zone entries.  If they create more, then they’re performing better than average in the offensive zone – if they create less, they’re below average.  Thus we can create an offensive zone score – basically showing the player/team’s fenwick they would have with their offensive zone results if their neutral and defensive zone performance was perfectly average.  We can do the same thing for the defensive zone (the exact opposite).

Fourth:  The creation of Neutral Zone Score, Offensive Zone Score, and Defensive Zone Score led to the most interesting findings about the neutral zone:  For individual players, only performance in the neutral zone was repeatable over a half season sample!  In other words, if you knew a player’s scores in all three zones over a period of time, you’d expect his offensive and defensive zone #s to regress heavily toward average as more time goes on, while his neutral zone #s would at least somewhat remain similar even as more games would be played.

For teams, later research suggested this was not the case: Team Performance in the offensive and defensive zone over a half season is in fact repeatable.  Neutral Zone Performance is still the most repeatable of performances, but unlike individual players, teams could repeatably thrive or fail based upon offensive and defensive zone performance, and some teams certainly do so.

In Sum: This suggests that the neutral zone is an extremely important feature of an NHL game and that stats measuring it are similarly important.  Teams that win the neutral zone may be the best at repeatedly winning the possession battle, which of course means they more often win the actual battle of winning games.  Players who win the neutral zone battle are more likely to put up good possession #s repeatedly than players who don’t, because performance in the other two zones doesn’t seem to be repeatable.  This means of course, that if a player has good or bad possession #s stemming from performance in the offensive and/or defensive zones, you can expect regression to pull him back toward average (or back toward his neutral zone average).

This means that neutral zone tracking is especially important!  Unfortunately, NZ Tracking has kind of fallen out of favor due to a lack of guides of how to actually do it.  So later this week, in Part 2, we’ll go over how anyone interested in neutral zone tracking can do it themselves to find out how their favorite team or player is doing in the neutral zone.


Leave a Reply

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

WordPress.com Logo

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

Facebook photo

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

Connecting to %s