Over the last few years, I've spent a considerable amount of time reading, watching and writing about hockey. I've played and watched the game for the majority of my life, but the way I did those things changed when I started doing what I do now.
Instead of watching a game and deciding whether a player was good or bad, I started to focus on a player's value in the context of what purpose he serves and in what situations he might benefit a specific team. In essence, isolating his skillset to determine his role. This isn't groundbreaking stuff, obviously, but it wasn't something I previously did when I focused on goals, assists, wins and losses.
The thing that I've been thinking about lately is a team's ability to move the puck. Possession is a concept that has been discussed with more frequency and an emphasis on maintaining puck control has become paramount in the hockey world. But how one grasps possession is what's been bugging me.
Hockey has seen an increase in quantifiable data that has attempted to provide a more defined view of the game. Shot differentials, zone starts and entries have all been crafted as ways of going beyond the traditional box score scouting guide. But what's eluded me is a way to quantify a player's positioning and puck distribution skills, two components that I believe are crucial to maintaining puck possession.
How does one go about determining these things?
@mattbrigidi watch the game, nerd.— mikeopinions (@mikeopinions) October 25, 2013
Yes, you could watch the game. But when, say, 10 games are being played (as there were on Thursday night), watching all the action is difficult to track. More finitely, there are 10 players competing on a single shift, that tracking player movement on that micro of a level is difficult enough in its own right.
For example: the Philadelphia Flyers dressed Jay Rosehill in an exhibition game. Most people think of Rosehill as a bum who only throws his fists. Well, in this particular contest I believed Rosehill exhibited excellent positioning on the back check and displayed discipline on defense that aided in the team's ability to consistently clear the zone.
Short of typing those words in an attempt to communicate that belief, I have no way of proving it. Yet, his positioning could be crucial in developing a playing relationship with other members of the team. This is especially true in hockey where players transition from playing defense to offense, back to defense, back to offense in a matter of seconds. A player's ability to move the puck and receive the puck is the difference between a successful breakout or an extended zone stay. In addition, his ability to grasp the concept of proper positioning in order to aide in transition is another key component in establishing chemistry.
But what is chemistry? In the world of narratives and eye tests, chemistry seems like a throw away line playing into a generalized thought. We have ways of inferring into the way a group of players work together but the picture is still a bit grainy. While we can say a particular line combination generates a great deal of success, it still remains unclear why they are generating that kind of success.
Also, when it comes time for free agency frenzies and trade deadline debates, establishing a player's value in relation to a potential suitor is still somewhat lacking. A mid-level acquisition who is "gritty" or "the glue" might have a tangible skillset that makes him "the glue," beyond the things we don't have ways of explaining.
One exciting way that this type of data could be generated is the SportsVU cameras developed by STATS LLC, which track real-time movement in an attempt to gain a deeper understanding of how games are played. These cameras have been installed in every NBA arena and will be used by every team in the NBA this upcoming season.
The Toronto Raptors have already been using this data and have developed a code for comparing where their players stood on a particular play and where they should have been. It looks like this:
A more in-depth explanation of this system can be found in this feature by Zach Lowe at Grantland.
As of this writing, the NHL hasn't indicated that they plan to use this technology. Based on the rapid growth of information, I'd imagine teams will be looking to use it, or something like it, in the very near future. That is, if they don't have their own system already in place.