Zone Entries: Introduction to a unique tracking project

Eric T.
June 20 2012 11:03AM

About the project

For the last year, Geoff Detweiler has been tracking the Flyers' zone entries and I have been writing about them for Broad Street Hockey. This article will be the first in a series in which I hope to introduce a new audience to the work and inspire a few readers to try tracking other teams.

The basic idea is pretty simple: by just recording when the puck enters the blueline from center ice, we can work out how each team does in the offensive zone (by how many shots/scoring chances/goals they get per offensive zone possession), in the neutral zone (by how often they push the puck forwards into their offensive end), and in the defensive zone (by how many shots/scoring chances/goals they give up per offensive zone possession).

We can look at how the team does with a certain player on the ice, or with a certain player carrying the puck. We can determine the amount of offense generated by carrying the puck in or dumping it in, and consider the defensive risk associated with each. We can look at how teams adapt their strategy to the score. We can look at how much impact a puck-handling goalie has on the opponents' attempts to play dump-and-chase.

The Critical Neutral Zone?

After a year of tracking, we've found some really surprising things. It looks like dumping the puck is a big tactical error if there is any other option, for example. We found what looks like an inefficiency in the Flyers' lines, with a player who has pretty mediocre results in the neutral zone handling the puck a lot more than he should.

But perhaps most surprisingly, the initial returns from this effort suggest that neutral zone play may be much more important than had been previously realized - we found essentially no difference between Claude Giroux and Zac Rinaldo in number of shots generated in the offensive end or number of shots prevented at the defensive end. A player's Fenwick score (the team's shot differential with them on the ice) seemed to be largely determined by their performance in the neutral zone.

This result is very surprising, and should be met with a critical eye; extraordinary claims require extraordinary proof, and right now we just have one team's data from one season. Later in this series, we will present the data that led to that suggestion, but for now I will begin with a call to arms: if you have a Gamecenter Live subscription and nothing better to do over the summer, think about the fun of making some unique discoveries about your favorite team and helping to change our understanding of the game at the same time.

How to track zone entries

The mechanics are relatively simple. All you have to do is watch the game and each time a team sends the puck into their offensive zone, you note which player did it, the time on the game clock, and a one-letter code for how they sent it in (D = dump in, C = carry in, etc.). You end up with a spreadsheet that looks something like this: 

Period Time Entry type Player
1 19:55 D Opp
1 19:28 C 17
1 19:12 D Opp
1 19:00 D 19

We chose not to note the individual opponents to simplify tracking - it was harder to instantly identify opponents, and we wanted to minimize the amount of rewinding required so that tracking a game wouldn't take any longer than watching it normally would. We excluded dump-and-change plays where the team makes no real effort to recover the puck. We also added a column for labeling which entries were the product of odd-man rushes, but most of the analysis can be carried out without that.

Then we copy-and-paste the NHL play-by-play into the spreadsheet and run a macro that does all the analysis. The macro divides the game up into segments that begin with a zone entry or faceoff and end with a clear or a stoppage and calculates how many shots were taken in each segment. Then it spits out a bunch of statistics that tell us how often a given player is involved in bringing the puck into the offensive zone and how effective they are at it.

The macro also spits out URLs that can be copied and pasted into a browser to use Vic Ferrari's scoring chance app to extract information about who was on the ice for each of those entries, which allows us to go beyond the individual puck-handling and analyze how effective the team is in each zone with a given player on the ice.

That's all it takes. If you want to give it a try, contact me by email (BSH.EricT -at- gmail) or on Twitter (@BSH_EricT) and I'll forward and explain my template spreadsheets, and I'd be happy to help with any questions you might have about processing or interpreting the data.

Previously by Eric T.

 

2654ef2681c88bc3252431ec45e30590
Eric T. writes for NHL Numbers and Broad Street Hockey. His work generally focuses on analytical investigations and covers all phases of the game. You can find him on Twitter as @BSH_EricT.
Avatar
#1 Corey S.
June 20 2012, 12:50PM
Trash it!
0
trashes
+1
0
props

This is one of the many things I will be working on in the off-season.

Avatar
#2 David Johnson
June 20 2012, 01:39PM
Trash it!
0
trashes
+1
0
props

"But perhaps most surprisingly, the initial returns from this effort suggest that neutral zone play may be much more important than had been previously realized - we found essentially no difference between Claude Giroux and Zac Rinaldo in number of shots generated in the offensive end or number of shots prevented at the defensive end. A player's Fenwick score (the team's shot differential with them on the ice) seemed to be largely determined by their performance in the neutral zone."

I am not sure this is completely surprising. Let's consider a perfectly played game where a team takes a shot, loses control of the puck, the other teams skates the puck down the ice, takes a shot, loses control of the puck, and they continue alternating shots in this fashion. Under this scenario each teams fenwick will be neutral.

Now how can that perfect shot alternating game get messed up so that one team benefits in the form of a positive fenwick?

1. A team collects rebounds for secondary shot attempts.

2. A team wins face offs to gain control of the puck more than their opponent.

3. Turnovers (where puck gets transferred from one teams control to the other via turnover, giveaway, dump in, etc.).

Some players are probably better are retrieving rebounds than others based on their positioning but I suspect rebounds are somewhat random as well (luck of the bounce) so I am not sure rebounds are a significant factor.

As for #2, only 3 teams had a face off percentage higher than 52% this past season (Boston, San Jose, Vancouver) and only 4 teams were below 48% (Edmonton, Anaheim, New Jersey and Calgary) so very few teams can dominate fenwick due to face offs.

That leaves turnovers where a team gets control of the puck but doesn't generate a shot attempt. Turnovers probably account for the majority of the difference in teams fenwick scores. If we tracked turnovers (puck transitions from team to team) I suspect the majority of them come from dump ins or when a team fails to maintain control of the puck while attempting to skate/pass the puck into the offensive zone.

So, long story short, I am not completely surprised by your observation.

Avatar
#4 SmellOfVictory
June 20 2012, 05:15PM
Trash it!
1
trashes
+1
0
props

I know there's been a ton of work done on Fenwick and scoring chance correlation on an overarching basis, but for this set of data, did you also take the proportion of scoring chances created for each entry by the players? A terrible player can gain entry and take an outside shot, and it doesn't mean much of anything. I'm curious as to whether there are any differences or not.

Avatar
#6 gongshow
June 21 2012, 11:06PM
Trash it!
0
trashes
+1
0
props

Thanks Eric. I've been hoping for someone to start this analysis for the last few years.

Comments are closed for this article.