Advertisement

Select Page

Quantifying The Value of Pre-Shot Movement

Quantifying The Value of Pre-Shot Movement

Despite the many new metrics being developed to evaluate goaltending, much of what happens to goaltenders on the ice is still hidden from statistical view. This puts limits on what these statistics can really tell us about goaltending or goaltenders. So, over the past few months, the InGoal Magazine staff reviewed over 2200 regular season goals allowed by 17 starting goaltenders in 2016-17 to identify trends and tendencies, for both individuals and the group.[1]

What we found matches what other trackers have noted: that plays crossing the center of the ice have enormous impact in the NHL.

It’s not possible to use this data as it stands to determine whether a goalie is “good” or not compared to other goalies. For one thing, it doesn’t include information about all shots faced. (Shot tracking is highly resource and labor intensive for a project this size.) For another, goalies and teams fit together in unique ways. What might be a weakness on one team could be a minor quirk on another.

We can use this database, however, as a window on what kinds of plays tend to trouble goaltenders the most, and what events should be incorporated into goaltending statistics going forward.

In our tracking, cross-ice puck movement played a role in 53% of all goals, more than any other element tracked. In other terms, a goal in the NHL is slightly more likely than not to include cross-ice puck movement. These goals are ubiquitous.

No other element tracked even came close. The next two most common elements were rush shots and quick releases such as one-timers, each of which occurred in 38% of all goals.

% of all goals with
Screen 23.1%
Deflection 21.1%
Rebound 21.4%
Breakaway 10.2%
Low-High 18.4%
Quick Release 37.6%
Against Grain 13.0%
High Lateral 22.5%
Cross Slot 30.8%
All Cross Ice 53.4%
Rush 38.4%
Power Play 21.1%

More than this, lateral plays dominated goal scoring regardless of what else was happening on the ice. Lateral puck movement occurred in 35% of goals with screens, 44% of goals with deflections, 43% of goals with rebounds, and 54% of rush goals. In virtually every case, lateral movement was the most common additional factor in these categories. [The exceptions are a very close relationship between breakaways and rush shots and between low-to-high movement and quick releases. In both of these, cross-ice movement is the second most common additional factor.]

Not surprisingly, cross ice goals often also had a quick release element — 53% of all lateral play goals included a quick release. The proportion was slightly higher for plays that crossed the slot below the top of the faceoff circles (56%) than ones that crossed the ice further up (49%), but not by much.

Importantly, 76% of quick release goals contained a cross ice element, suggesting that lateral movement is the more dominant force in this relationship.

In other words, lateral play goals occur more often without quick releases than quick release goals occur without lateral movement. It seems that cross ice puck movement does more to either create or intensify the danger of quick releases than vice versa. This is borne out in the raw numbers as well: 851 goals included a quick release, while 1208 included lateral puck movement.

There simply is no other element that makes as much noise as cross-ice puck movement prior to a shot.

We further broke down scenarios with lateral movement into plays where the puck crossed the center line below the top of the faceoff circles (cross-slot) and where they crossed the line higher in the zone (high lateral). Plays that crossed the ice more than once, both above the circles and below, were counted as cross-slot plays. Cross-slot movement occurred on 58% of all goals with a lateral component and 31% of all goals tracked.

Largely, the two had similar profiles, occurring with most elements between 12 and 18% of the time. However, high laterals were far more likely to include a screen or deflection, which makes sense considering that there is more opportunity for someone to get in between the goal and the shooter. On the other hand, cross-slot goals were more likely to come on a rush, where the defense did not have time to get set after entering the zone.

In essence, this data indicates that at the NHL level, taking advantage of a goalie’s inability to cross lateral space quickly (or as quickly as a puck can) is somewhat more valuable than trying to disrupt the goalie’s visual attachment.

While this data can’t shed much light on the efficiency of cross-ice puck movement, the sheer dominance of lateral plays among a wide range of goal-scoring scenarios and a good cross-section of the league’s most active goaltenders, suggests that it could be high. At the very least it’s something worth looking at further.

The difficulty is that most of the widely available metrics for goaltending rely on data gathered in the RTSS system by the NHL’s off-ice officials and published in the official play-by-play files. This data doesn’t identify pre-shot movement or most other scenarios that complicate making saves, and in most cases (rushes and rebounds being partial exceptions) no one has yet discovered a method of gleaning that information from the files.

That leaves manual tracking of these goaltending elements as the only current method for seeing them. While this is time-consuming, the potential value for goaltender evaluation could be enormous.

One final but very important caveat: this is a single sample of goals in the NHL, so it is important to be careful about drawing strong conclusions. All data gathering has the potential to include some observer bias, and this is no exception. As we track more goals in the future, the picture could change. This should be regarded as a preliminary study, one that can help us formulate better questions to ask about goaltending and about the ways we could improve statistical evaluation of goaltending.

 

[1] The goalies included were starting goalies in playoff teams: Jake Allen, Frederik Andersen, Craig Anderson, Sergei Bobrovsky, Corey Crawford, Devan Dubnyk, Brian Elliott, Marc-Andre Fleury, John Gibson, Braden Holtby, Martin Jones, Henrik Lundqvist, Matt Murray, Carey Price, Tuukka Rask, Pekka Rinne, and Cam Talbot.

About The Author

Clare Austin

Clare Austin is a reluctant "stats nerd" living in Nashville, where she has never worn a cowboy hat or boots.

Leave a reply

Your email address will not be published. Required fields are marked *

InGoal Partners

Follow InGoal on Social

ingoalmag
6 days ago
Andrei Vasilevskiy w

559

1

Andrei Vasilevskiy was the game’s first star and Jake Allen kept his team in it as the Tampa Bay Lightning beat the St. Louis Blues 2-1. Take a minute to enjoy some of their best work in tonight’s Save Story.

Read InGoal Magazine