clock menu more-arrow no yes

Filed under:

It's All Relative: Introduction to Neutral Zone Scores For NHL Defensemen

New, comments

A look at defensemen and neutral-zone play around the NHL

Kevin Hoffman-USA TODAY Sports

[Note: The majority of the legwork for this post was done by Corey Sznajder, who is manually tracking all 1230 NHL games from last season. Read more about Corey here, and support his work here. Special thanks to our own Muneeb Alam for his work on retrieving the on-ice data used in this piece]

We have written a lot about neutral zone play this summer, from Barry Trotz and his potential impact on zone entries, to the Capitals' top trio, and most recently about offensive contributions.

The majority of our zone-entry analysis up to this point has been centered around Capitals players (for obvious reasons). Now thanks to some recent data provided by Corey Sznajder, and the skills of Muneeb Alam, we have on-ice zone entry data for all 30 NHL teams - which will let us get an idea of how defensemen are performing in the neutral zone around the League.

The best current metric for accessing neutral zone play is "neutral zone score" (NZ score), which was first presented at the 2013 MIT Sloan Sports Analytics Conference by Eric Tulsky, Geoffrey Detweiler, Robert Spencer, and Corey Sznajder.

Nz_score_explanation

"Here, Sh_EWP= shots per entry with possession, Sh_EWoP= shots per entry without possession, EWP= entries with possession, EWoP=entries without possession...an O at the start of a terms means it is the result allowed to the team's opponents, and an asterisk superscript means the league average for that metric...the neutral zone score is a function of how many of each type of entry the team gets and allows" (Tulsky, Detweiler, Spencer, and Sznajder).

Put another way, NZ score is essentially what a player's Fenwick For% would be if Fenwick attempts from faceoffs were not included, and each controlled and uncontrolled entry was worth the league average Fenwick value for that type of entry. Because NZ score (like Corsi For% and Fenwick For%) is derived from on-ice data, it has limited value when comparing players from different teams. To allow for better player to player analysis a "NZ score relative" metric can be used.

NZ Score relative is determined by comparing the team's NZ score when the player is on the ice, to the team's NZ score when the player is not on the ice. It is a measure of how the team performs in the neutral zone with a given player on the ice relative to how they do without him.

The following table shows all NHL defensemen that had their zone entries tracked for a minimum of 250 5v5 minutes, ranked by their NZ score relative:

Click on any column header to sort by that column

While looking at this data, it's important to note that while neutral-zone performance is a huge component of puck possession, it's not all-encompassing. There are a few teams that struggle in the neutral zone but make up for it with their offensive and defensive zone play, and vice versa but that's a topic for another day.

Let's narrow our focus to just players under contract with the Caps for the 2014-15 season:

It is probably a little hard to believe that the Capitals had two defensemen in the top-three of any category last season, let alone one that is associated with puck possession. But while Mike Green (whose statistical value we already examined) and Dmitry Orlov (who we have ranked as the Capitals' 3rd best player under the age of 25) may give up a lot of entries with control, they more than make up for it with their work in the neutral zone... in part due to the sheer volume of entries, with both Green and Orlov ranked in the top-20 of eligible defensemen in terms of % of On-Ice Z.E For. If Orlov's arm is healthy in time, a return to the Orlov-Green defensive pair in the upcoming season might not be such a bad thing. But you knew that already.

Meanwhile John Carlson's 53.2% and Karl Alzner's 53.6% were the fourth and second highest control percentage-against among eligible defensemen. It is unclear as to whether Alzner and Carlson will find themselves playing together this season; if they do, though, it will be interesting to see what sort of impact a new system has on their NZ numbers. As for the rest of the carry-overs from last year's roster, it's hard to imagine Steve Oleksy, John Erskine, or Nate Schmidt seeing a lot of playing time if Orlov is healthy. Schmidt's NZ numbers in particular are a little surprising, as he had good possession numbers last year (especially when paired with Mike Green), but that does not seem to be reflected here.

So what about the Caps' newest blueliners? Matt Niskanen had a good NZ score relative while Brooks Orpik had a bad one, but that doesn't tell the whole story. Niskanen's opponents only entered the offensive zone with control 39.8% of the time, one of the lowest percentages in the entire NHL, but Orpik's Control% Against was low as well, perhaps providing some more support to a theory that was floated shortly after his signing. Both players are likely to be positive additions.

The fact that every single Capitals' defensemen had a higher control against% than control for% last season is troubling, but not surprising, and a big part of the problem was the team's poor neutral zone defensive strategy overall in 2013-14. The addition of Niskanen and Orpik (yes Orpik), together with the systematic changes likely to be implemented by Todd Reirden and Barry Trotz, make it not only possible but probable that the Capitals' neutral-zone play shows marked improvement in the upcoming season.