Intro

Hockey analysts have repeatedly shown the value of neutral zone play. If a player performs well in the neutral zone, he or she is helping generate offense for their team and limiting the opponent’s chances. In addition, neutral zone play is repeatable, and the player is likely to continue to drive possession for their team. If you can identify players who thrive in the neutral zone, you are in a position to help your team improve.

But while neutral zone play is important, we still have a very limited understanding of it. Between the distance from the goal, the fluidity of play, and the relative scarcity of data, most people don’t know how players perform in the middle third of the ice. Furthermore, we don’t even have a complete idea of how to make those evaluations. When figuring out how good a player is in the neutral zone, should offense and defense be evaluated separately, or are overall results enough? What skills translate to strong neutral zone play? What playing styles?

While I cannot answer all of those questions, this post addresses the first one by investigating the relationship between getting into the offensive zone and preventing the opponent from doing the same. I analyze entries for and entries against to show that gaining zone entries and preventing them are two unrelated skills. This material is based on the first half of the presentation I gave at the Vancouver Hockey Analytics Conference, and like many recent advances in public neutral zone knowledge, is based on Corey Sznajder’s work tracking every game of the 2013-2014 season.

Methods and Hypotheses

Is the best defense a good offense? Micah Blake McCurdy’s article here showed that “shot generation and shot suppression are mostly independent”. In other words, the players that generate the most shots for their team do not necessarily prevent shots against. But what about in the neutral zone? Are players who can skillfully enter the zone also able to deny their opponents? There are three possibilities:

1. There is a direct relationship between entries for and entries against. While intuitively unlikely, this would mean that players tend to stick to a certain “pace”: some see a ton of back-and-forth entries for both teams, while others slow down play and limit movement on both sides.
2. There is an indirect relationship between entries for and entries against. This implies that when the puck is in the neutral zone, both teams battle to turn it into an entry, and gaining an entry for your team denies an entry to your opponent. This is “the best defense is a good offense”.
3. There is no relationship between entries for and against. A player’s ability to get zone entries is separate from his ability to prevent opponent’s entries.

Results: The best offense is just a good offense, the best defense is just a good defense

I used methods similar to those in McCurdy’s article. I looked at a dataset of 5v5 entries for per sixty minutes (EF/60) and entries against per sixty minutes (EA/60) for every game from every player. To remove some outliers, I only looked at games where the player had at least 7 minutes of TOI. This left me with 40,000 data points that produce this scatter plot:

Note that a few points are cut off from the right side of this chart. The striation is fun but not important here, and I’ve made each point very transparent, so the darker places are where more data points fall. At first glance, there is no obvious pattern.

However, this visual is not very clear. Moreover, it does not account for differences in ice time: 30 minutes of Erik Karlsson is equal to 8 minutes of Tanner Glass. (This post is the only time that that sentence will ever be true.) I’ve applied similar methods to McCurdy’s work to address both. First, the scatterplot has been replaced by a density curve, which is essentially a contour map showing the general shape of the data. Second, I’ve updated the data so that each point is weighted based on the player’s TOI.

If there was a clear relationship, these lines would be diagonal ovals. Instead, they are basically circles, indicating that there is no relationship between generating entries for and preventing entries against. This is further supported by the R² value of 0.012.

Generating and preventing entries are two separate skillsets in the neutral zone, and looking at each one separately can help understand how a player produces results. This remains true when breaking the data down by position between forwards and defensemen or by entry type between carries and dumps.

A player who is great at entering the zone may or may not be just at good at stopping opponents. Just as Corsi For % masks the finer details of generating and suppressing shots, overall neutral zone scores can give further insight when broken into their offensive and defensive components. For example, Gabriel Landeskog and Marian Gaborik each have EF/60 near the league median (73), but Gaborik allows about 16 more entries against per 60 (84 to 68). Looking at both sides can help front offices evaluate players and coaches determine what skillsets need attention.