One of the incoming changes to the National Hockey League for the 2019-20 season is the implementation of puck and player tracking technology in all 31 arenas. This was hyped up during the 2019 All-Star Weekend and it was performed in a few games last season, as per this January 25, 2019 NHL.com article by Tom Gulitti. From the article, here is a summary of how it will work:
The NHL Puck and Player Tracking technology will include 14-16 antennae installed in the arena rafters; four cameras to support the tracking functionality; one sensor placed on the shoulder pads of every player on each team; and 40 pucks manufactured with a sensor inside for each game. The technology is to be deployed in all 31 NHL arenas at some point during the 2019-20 season.
The hope is that the flood of new data will enhance the fan experience inside the arena and for those watching on television, via mobile devices or, in some cases, both with a digital stream broadcast dedicated to showcasing puck and player data.
Essentially, the technology will be able to output thousands of events that otherwise were not recorded. In addition to identifying where a player (or a puck) is for a shot, the technology would identify where the other nine skaters were for that shot. In addition to noting when a hit is thrown, the technology can identify the location of the hit and how many were involved - and whether the puck was anywhere nearby. The technology can track where passes - successful and unsuccessful - go as well as when zone entries and exits are completed. More than how long has a player skated, the data may be able to map out their pathways.
Gulitti’s article and others like it focus on the entertainment aspect. I am more interested in the informational aspect. I believe most teams are and those that are not should be. Any team with an analytics department should be prepared for a massive influx of data and should have a team ready to not only parse the data technically but also holistically. The Devils hired Tyler Dellow and Matt Cane earlier this year and I would think they both understand what is coming down the proverbial pike. Whichever organization can get additional value out of the data will have an early advantage. For a Devils team that is looking to compete sooner rather than later, any advantage is desirable.
I have no idea how much or how little the new tracking information will be made public. For all I know, none of it will be until long after NHL teams recognize it, utilize it, and do not see any disparity in stating it like they did with shot attempts. Nonetheless, I want to blue-sky some possibilities of what puck and player tracking can bring to the table that we never thought before.
How often have you or some one else said, “They have to get traffic in front?” How often have you seen a play for a long-distance shot where someone on the same team is or tries to get in front of the goalie? I would say every team in the league utilizes screens in someway or form. Some are intentional like on a power play. Some are ad hoc in the run of play. We know some players are better at this than others. For example, Patrick Maroon was quite good at this for the Devils in 2018. But I admit there is little evidence to justify that outside of spot examples and the old eye test. With puck and player tracking, this evidence would be possible. Just by knowing where the other players are on the ice for an shot attempt, we would know:
- Is there a screen?
- Who is setting it?
- Are there other defenders there?
- And based on the shot itself, was the shot was on target or not with the screen?
- Depending on how precise the tracking is, if the shot did not go in, did it hit the screen instead of the goalie?
We see screens utilized regularly. For that alone, we should be able to answer the question of whether screens are a good idea. Tracking will be able to do this. We may assume that larger players would do well at setting screens. Tracking will be able to determine if that is true or not. This is just one aspect of the game that currently is not really being measured that tracking can make possible.
And this would also add another level of analysis for goaltenders. It is currently basic: how many shots were stopped as if all shots were the same. It is harder for a goalie to stop something they cannot clearly see. Even just knowing if a shot against had a screen or not would at least let teams breakdown save percentages that way. Instead of punishing goalies and save percentages for having traffic in front, something they cannot control, this kind of data would really add more nuance to a position to badly needs it from an analytical perspective.
Here is another part of the game that we see regularly but we have no real information about. Players engage with each other along the boards in the hopes of obtaining possession of the puck. Many players, particularly larger defensemen, get credit for being “good along the boards,” but again we do not have much to back it up. More importantly, does winning the puck along the boards mean anything on its own? Or is it only valuable if it leads to a shot attempt (offense) or a zone exit (defense)? This is something that could be observed today but with TV cameras only showing so much and the puck often being caught under skates and sticks, it is not always easy to figure out who won. Much less determine whether it is better to be first, bigger, stronger, etc. in a battle. With puck and player tracking, we can start getting some answers.
Again, assuming that the puck’s location is tracked during the game, one could define and identify when a battle for possession takes place. It may also be possible to identify who is involved - including position - and where it is at. That alone would be large. It would provide some evidence as to whether it is the size of the dog in the fight or the size of the fight in the dog. There might be more risk than not to do it in the corners than along the sideboards, for example. There may be a possibility to establish whether it is worth going over to help a teammate to with such a battle or not. Similarly, the data may be able to sort out whether a defender should force such a battle or if they should focus more on what the other players are doing at the time.
Dumping the puck is not an efficient way of gaining the zone. This has been known for years. Yet, lines are instructed to do so and players do it anyway. It may not be a good decision (especially on the power play) but if that is what the coaching staff wants, then we should be able to know whether it is succeeding or not. Or at the least provide the data that shows that dumping-and-chasing may not be so efficient. Tracking data can accomplish that and reveal more about the dump-in.
Through puck and player tracking, it may be possible to identify where the dump-in takes place, where the dumped puck actually goes (or how deep into the zone does it have to go for it to be deep enough), who retrieves it, and even the time between the dumping and the retrieval. That last possibility is important to understand whether it is effective to perform a dump-and-change; what kind of dump-in has to be performed to allow for enough time for players to change. Similarly, the data may reveal who is more effective at performing the dumping, which teams and players are effective at cutting it short, and what kinds of players would be best for it. Miles Wood of the Devils is seemingly involved in a lot of this. I do not like it, but the data can prove whether or not it makes sense for the coaches to use him this way.
Defenses do not have a lot of stats actually associated with the concept of playing defense. So much of it in hockey is based on positioning and reacting appropriately to plays as they happen. With puck and player tracking, teams may be able to get a legitimate handle on gap control. It may vary from player to player, but the idea is for the defending player to stick close enough to an attacking player to keep them at bay but not so close that the attacking player can get around him easily. It is a concept that may really apply for plays off the rush both even situations (1v1, 2v2) and odd-man rush situations. It is still something that is specific to the defender where a team can analyze how close is too close, how far is too far, and what an ideal length would be for most plays. It alone would be a step forward in terms of identifying who does well in such a defensive situation.
If the player tracking data is precise enough to identify where a player is facing, then that would be a huge benefit for analyzing defense. There is a big difference between being a foot from an attacking player and facing him or being a foot from an attacking player and facing in a different direction unaware of him. The former has a chance of making a defensive play or take that player out of the equation. The latter is in trouble. I do not know if tracking will go that deep but at least knowing where other players are on defense in relation to where the attackers are would be a step forward.
Zone Entry & Exit Details
While Corey Sznajder has performed a lot of hard work on his own for tracking zone entries, zone exits, and passes that lead to shots, the tracking has focused on who performs the entry or exit and how it was done. That’s great. The puck and player tracking should be able to do that as well. It should also be able to identify certain structures and possibly common traits about them. Such as how many teammates are near the player when they make the entry. Or how many touches or passes are made before an entry or an exit. Or how much time it takes to go from a zone exit to a zone entry. Or whether transitional play happens more when the other team is caught with too many bodies in the offensive zone. Or whether certain players are present on the ice for more shooting attempts, zone exits, zone entries are doing anything in particular that is noteworthy.
The work that Sznajder has performed is admirable and has opened many eyes to who is truly creating offense. Puck and player tracking can take it to the next step and reveal how much other players are involved. Especially on power play situations; one player may be carrying it in but how involved are his four other teammates? Tracking can identify that those not directly involved are providing value just by making the defense respect them.
Tips & Deflections
Currently, a tipped or deflected shot is recorded if it leads to a goal or a shot. Similar to screens, it is something that skaters try to do and goalies loathe. But we have no evidence as to whether anyone is particularly good at it outside of whether they score on it. The concept, like a screen, raises more questions about what happens. What about all of those other deflections or tips or inadvertent touches of a shot puck that can make the difference between glory and defeat? What about the goalie, whose save percentage just considers all shots to be shots taken at them instead of separating out deflected shots? What about a deflection that re-directs a shot away but for one reason or another is not called a blocked shot? Depending on its precision and resolution, the puck and player tracking should be able to answer all of this and more.
A lot of hockey analytics is based on what is recorded. But what is not being recorded could have an influence on the game as a whole. Like a metagame, there may be actions that we know and recognize in games but do not fully understand why they take place. I’m very excited to see how the game may change with all of this additional puck and player tracking coming into place next season. And I’m hopeful that Dellow, Cane, and the rest of the staff that would work on this will come to some actionable conclusions sooner than most other teams for a better, more efficient, and more effective Devils team. The team has more talent, but utilization and instruction of that talent can possibly make a big difference in terms of success. Analysis of data is one way to get there and there’s going to be a lot of new data coming in soon.
These are just some of the possibilities I can think of off of the top of my head that puck and player tracking may provide. What do you think such data may reveal about the game we all love? What gaps will it help fill in for what is already tracked or collected about hockey? What areas of the game do you think this data will help out in? How do you think it should be used in broadcast and studio segments about the game? Should it ever be public? Please leave your thoughts and other blue-sky thoughts about player and puck tracking for next season in the comments. Thank you for reading.