Photo credit: NHL.com
This past weekend I was given the opportunity to speak at the Rochester Institute of Technology Hockey Analytics Conference. It was an honor to be part of a wonderful event with such smart and talented individuals. You can find a recap with links to videos and slides of each speaker at Hockey Graphs. On a personal level, this was really difficult for me. I’m a fairly reserved person and I generally don’t talk much around people, even my friends. Naturally, I kind of freaked out a bit when it came time to present (ok I completely freaked out). I have no idea whether or not my ideas came across well at all so I wanted to take the opportunity here to breakdown some of what I attempted to discuss this past Saturday. Hopefully this will give a better idea of the changes to zone exit tracking I plan on implementing this season and what we can hope to learn from them.
Breaking Down Pressure Situations
The image above is just a rough idea of the pressure areas a player may face when trying to exit the zone. Please feel free to laugh at my poor MS Paint skills.
When I first started tracking zone exits it was pretty straightforward; note how player x gets the puck out of the zone, end of story. While the simple method may sometimes be best, I felt that something was lacking in the way I was tracking zone exits. At one point I started to get the impression that forwards generally had an easier time exiting the zone than the defense. If that is in fact the case then how impressed should we be by a forward that has a high success rate at exiting the zone, even with possession? I wanted to find a way to incorporate some tracking parameter that would shed some light on this and determine whether or this was true or was my memory deceiving me? Spoiler alert: my memory was terrible.
Last year I added a simple “yes/no” pressure category at the suggestion of Andrew D from Broad Street Hockey so we can try to get an idea of the difficulty players are facing when they attempt to exit the zone. While this helped add another layer of perspective to zone exits I felt that there had to be another way to improve upon how we view pressure. Throughout the course of a game players face varying degrees of pressure when they have the puck in the defensive zone. After reading the book Hockey Plays and Strategies, I decided to take their breakdown of different pressure situations and utilize them in exit tracking for this upcoming season.
Pressure will now be broken down into four separate categories. Please note that I was only able to track about thirty games using these new parameters. This is still very much a work in progress so there may be changes made as the season progresses.
Category One will be used to identify immediate forechecking pressure. Under this category players are generally trying to get the puck out of the zone with an opposing player right on their back which makes exit attempts extremely difficult. The player has very limited time to read and react to what’s going on around them. Most of these exit attempts resulted in players chipping the puck out of the zone or simply turning it over due to an inability to escape from the heavy pressure.
Category Two will be used to for close forechecking pressure. Within this category we should be able to get a sense of a player’s decision making and awareness in the defensive zone. The most common exit attempts under this category were chips and passes. When dealing with this type of pressure players typically have a little more time and space in their attempts. Quite often when a player chipped the puck out of the zone in this instance, it had more to do with poor decision making on their part. For example, they would go and retrieve the puck in the corner and instead of taking a look over their shoulders to get an idea of their surroundings, they assumed an opposing player was right on their back. Rather than making a pass to an open teammate, or carrying it themselves, they opted to take the “safe” route and chip the puck out of the zone.
Category Three represents when a player is facing a lax forecheck. We’ll typically see this when a team is resetting in the neutral zone after a line change or they’re sitting back to protect a lead. At times they’ll send one player in to passively pressure the puck carrier but there is no actual urgency to regain control of the puck. The exit attempts were pretty much split between carries and passes. Troubles within this category may be able to pinpoint a couple of different issues. We’ll either get more confirmation that a player lacks skill, or we may see that there may be issues with their breakout scheme.
Finally, Category Zero will be used when a player is facing no pressure from the opponent. For the most part these events happen immediately after the opposing teams make a line change and get caught in transition. There were times where a Flyers player, usually a forward, would end up alone near the blueline after the opposing team got drawn over to one side of the ice but it didn’t happen too often. Obviously, carries were the most common exit attempt among players when facing no pressure.
Now I know there may be some concern regarding the amount of subjectivity that may affect how these different situations are categorized. That was definitely something I was initially worried about when I started creating these parameters. When I really started to get into it, I didn’t find myself second guessing or having internal debates about how to categorize an exit attempt that often. There are always going to be some borderline plays throughout the course of a game that can go either either way. When those situations did happen as I was tracking I tried to give players the benefit of the doubt.
Analyzing Player Contributions In The Defensive Zone
There is still so much to do to try and find ways to improve the way we evaluate not only defensemen but defensive zone play as well. Another issue I had with the way I was tracking zone exits was missing out on which players are actively involved in exit attempts. For this year, I plan on tracking primary and secondary assists as well as putting some focus on line combinations and defensive pairings.
According to the thirty games worth of data I have, there were only about 1.7 players involved in each zone exit attempt in that data set. I tried to take a really simplistic approach to track the number of players involved in the opponents’ exits as well, but it became too time consuming for the limited time I had to work on collecting data. However, in those few games I was able to track they had about 2.1 players involved in their exit attempts. I know you may look at those numbers and then wonder why I would bother adding a secondary assist column if they are a pretty rare occurrence. I’m sort of curious to know if the number of players involved may be related to systems, poor support from teammates, score effects, etc. After a coaching change this offseason this year may be the best time to see if any of those could be a factor.
The main takeaway from adding even just a primary assist category is really to get a better picture on how players actions on the ice impact their teammates. Is a player putting his teammates in a good position to get the puck out of the zone and helping to create a successful transition? Could he part of the reason his teammates turnover numbers or uncontrolled attempts are so high? Now, is this method going to be perfect and capture everything? No, but I’m hoping it will capture enough to paint a decent picture.
I’m just going to quickly touch on other changes I planned on implementing for this year rather than torture you any longer. Transition data will no longer be treated as a separate entity. They will be incorporated in my zone exit data and posts this year. I will also be taking a look at score effects on zone exit attempts and pressure situations.
Again, this a work in progress. There are always going to be some limitations to how much and what we can truly capture but hopefully this will help add another layer of perspective to what is happening the defensive zone. I’d be happy to hear and discuss any suggestions or ideas you may have so please feel free to leave a comment or reach out on twitter.