Oilers Microstat Tracking Glossary
An explanation into all the stats manually tracked in our microstat project
In the season of 2022-23, I began a manual tracking project to track various stats for the Edmonton Oilers, commonly referred to as microstats.
The type of statistics that the public analytics sphere predominantly uses are macro-level on-ice stats, such as goal differential (GF%) and expected goal differential (xGF%). They’re based on the NHL’s official PBP data, which provides information about the location of every shot attempt (distance, angle, etc). When used correctly, these stats are an incredibly useful descriptive tool, and they grant us a solid idea of a player’s on-ice impact on their team.
However, the primary flaw of on-ice proxies is that they’re unable to describe how a player achieves their results.
For instance, let’s say Player A has a 60 percent goal differential at 5v5, which is quite good. But how did they achieve those numbers? What are they specifically good at? How could they stylistically fit on a certain team, or with certain linemates?
Without video analysis, it’s difficult to accurately answer these questions with the data we have access to. In the grand scheme of things, the results matter most, but it’s important to recognize how a player or team achieved those results in the first place. To improve a player’s performance, or to improve a team, it’s important to know their exact strengths and weaknesses.
I often like to refer to this quote from Eric Tulsky, who’s currently the assistant general manager of the Carolina Hurricanes: “Public models generally rely on inferring a player’s impact from looking at how the team’s results change when he is on the ice – and that can be done very well, and can be powerful. But when you want to move from talking about what happened to talking about why it happened or what could have been done differently, you often need a richer data set.”
This is where microstats come into play. Microstats include a large variety of different statistics such as zone entries, zone exits, puck retrievals, and so on. With these stats, we can more easily identify the specific strengths and weaknesses of a player or team.
Unfortunately, the NHL’s official PBP data does not include this sort of data. These stats are typically only tracked by private companies and NHL teams, so official microstat data is not publicly available.
This is why I’ve decided to start this project. As an Oilers fan, I intend to solely track Oilers games for now.
If you haven’t already, I highly recommend checking out Corey Sznajder’s incredible work. Corey has been manually tracking microstats for over a decade now, tracking data for hundreds of NHL games each season. These stats are all available on his website AllThreeZones.
I also recommend checking out Cam Charron’s work. Previously working as an analyst for the Toronto Maple Leafs, Cam tracked microstats for the Leafs and Canucks in 2022-23, and was then recently hired as an analyst for the Pittsburgh Penguins this past August. You can view some of his work last year at The Athletic.
Essentially everything I know about microstats has come from these two talented individuals. Most of the stats tracked in this project are based on the stats they’ve tracked, alongside some of my own personal ideas.
In 2022-23, I ended up tracking 68 regular-season games for the Oilers, and all 12 playoff games. For the 2023-24 season, my current objective is to track all 82 regular season games, and every playoff game.
This time, I will have help from two other wonderful people with this tracking project: Shaedon (@Shae_Nuge93 on Twitter) and @papa_mcleod. They will be tracking the “Shots and Passing” section of this project, which will be explained in detail later on. I highly recommend giving them a follow.
This piece is intended to serve as a glossary that explains how each of the stats in the tracking project are defined. Without further ado, here it is:
Zone Entries
To win hockey games, you need to score goals. To score goals, you need to be able to enter the offensive zone, which is why zone entries are a major component of hockey. You’d be hard-pressed to find a single top offensive forward who doesn’t excel in this facet.
Zone entry attempts can be split into three categories: controlled, uncontrolled, and failed.
Controlled Entries are defined as entering the zone with control, by either carrying or passing the puck into the offensive zone. They can be split into carry entries (also known as carry-ins) and pass entries.
Uncontrolled Entries are defined as entering the zone without control. Typically, they just include dump-ins, which are classified as shooting the puck deep into the offensive zone to establish a forecheck. If a player simply chips the puck into the zone and goes for a line change and/or if no players enter the zone to recover the dump-in, the play will not be tracked as a zone entry.
Failed Entries are classified as simply failing to enter the zone with control or failing to dump the puck in.
Additionally, I track if a controlled entry directly led to a scoring chance (the chance must be generated within six seconds of the entry).
Now, defining scoring chances can be subjective, and can vary by tracker. To make it as simple and objective as possible, we define a chance as any unblocked shot attempt within the home plate area, which is the area highlighted in black in the visual above. This gives us an idea of the actual impact of a player’s zone entries, and if they can generate scoring opportunities off the rush.
Zone Exits
Zone exits are another crucial part of the game.
Breaking the puck out of your defensive zone is extremely important on both the offensive and defensive sides. An effective zone exit gets your team out of the DZ, and towards an opportunity to generate offence. Failing to get the puck out of the zone not only grants the opposition a chance to score, but it’s also a missed opportunity to move the puck up the ice.
Zone exits can also be split into three categories: controlled, uncontrolled, and failed. Icings will be tracked in their own separate category.
Controlled Exits can be defined as exiting the defensive zone with control, by either carrying or passing the puck out. They can be split into carry exits (also known as carry-outs) and pass exits (also known as pass-outs).
Uncontrolled Exits are defined as exiting the zone without possession, and they’re split into zone clears and missed exits. Zone clears are instances where a player intentionally dumps the puck out of the zone, whereas missed exits are defined as plays where a player attempts to make a controlled exit, but it doesn’t result in possession. For instance, if a defender tries to pass the puck out of the zone to a forward, but the pass doesn’t clearly connect, it’s tracked as a missed pass exit.
Zone exits are only awarded if at least one of the following two conditions are met:
One or more opposing forecheckers directly applied pressure on the player attempting to exit
Two or more opposing players were in the defensive zone at the time of the exit attempt
Exits under no pressure, most notably when the opposing team is off for a line change, are not tracked. I want these stats to be as reflective of a player’s ability as possible, and I don’t think there’s much skill involved in skating the puck out of the DZ while no opposition player attempts to pressure or regain the puck. Additionally, I typically don’t track zone exits after defencemen briefly retreat into their own zone, especially off a center-ice faceoff win, unless the defenceman is forced below the faceoff dots by opposition forecheckers (I based this criteria off the Tape To Tape tracking project from a couple of years back).
In my regular season tracking in 2022-23, I specifically tracked failed zone exits, which were essentially just defined as turnovers when attempting to exit the zone, but I modified this in the playoffs, instead tracking all general defensive zone turnovers. This method will be continued in 2023-24, and will be further explained later on.
Shots and Passes
This part is pretty straightforward. In a nutshell, we track shot attempts, shot assists, and additionally identify if the shot assist was a high-danger pass.
Shot Attempts are simply defined as plays where a player deliberately attempts to shoot the puck on the net. For now, we will track all unblocked shots. Shots are split into four categories; rush, forecheck, cycle, and faceoff shots. Here’s how they’re defined (these are based on Cam Charron’s definitions):
Rush Shots - Shots within six seconds of a controlled zone entry
Forecheck Shots - Shots within four seconds of a dump-in recovery or a turnover
Faceoff Shots - Shots within six seconds of a faceoff
Cycle Shots - Any other shots taken in the offensive zone that do not fit the above categories
This is the primary reason why shots are tracked, even though the NHL already tracks them, as we can split a player’s shots into these categories. It gives us more information on how a player generates their shots.
Primary Shot Assists are defined as passes leading to an unblocked shot attempt; they’re the same definition as regular assists tracked by the NHL, but for all (unblocked) shot attempts.
HD Passes include royal-road / cross-seam passes and passes from behind the net. Shots that come off these passes have a significantly higher shooting percentage than other shots, and due to the lack of pre-shot movement data publicly available, these sorts of shots can often be underrated by public expected goal models. Here’s a great article by Corey Sznajder that dives into how shots off these passes are so dangerous.
This is essentially the shot-tracking component of our project in a nutshell, but there are other minute details tracked about each shot, such as if the shot was screened, if the shot was a scoring chance (i.e. the shot was within the home plate area, as mentioned above in the zone entry section), if the shot came during an odd-man-rush, and so on.
Forechecking
One of the best ways to keep the puck away from the DZ is through forechecking and pinching. To evaluate forechecking, I track exit disruptions and dump-in recoveries.
Exit Disruptions are defined as forcing an opposition failed exit or icing. Players can disrupt exit attempts by pressuring an opposition player into turning the puck over, or by intercepting a pass attempt. Sometimes, two players can have an exit disruption on one play.
Dump-In Recoveries are defined as recovering a dump-in by beating an opposing defender to the puck. As for puck recoveries off a missed pass or a zone clear that goes into the offensive zone and not off an intentional dump-in, I keep track of them in a separate category, labeled as Other Loose Puck Recoveries.
Transitional Defence
A shot off the rush can be defined as a shot subsequent to a controlled entry. On average, a shot off the rush is often more dangerous than the average shot off the forecheck or cycle.
Consequently, defending the rush is a significant aspect of defensive play. The best way to do it is to prevent controlled zone entries, and many refer to this as zone entry defence.
Entry Targets are the total count of how many times an opposing player has attempted to enter the zone past a defenceman. The player must be in position to defend the blueline. No entry target is tracked on a two-on-one; instead, in most instances, a failed pinch will likely be assessed for the defender caught up ice. Plays where an opposing player dumps the puck in and goes for a line change are not classified as an entry target or an entry against.
Next, I track zone entries allowed by a defenceman, which is the total count of how often an opposing player enters the zone past the defender. They’re split into Controlled Entries Allowed and Uncontrolled Entries Allowed.
I also track Entry Denials, which are the total count of plays where a defender forces a failed (controlled) entry attempt from an opposing player.
The best entry defenders will deny entries at high rates, and allow more uncontrolled entries than controlled entries, as controlled entries are around three times more likely to result in a scoring chance.
I primarily track zone entry defence strictly for defencemen. However, if a forward forces a failed entry, I will assess an entry denial on the play.
I also track if a controlled entry against led to a scoring chance. Again, I define a chance as any unblocked shot attempt within the home plate area, and the chance must occur within six seconds of the entry. This gives us an idea of how well defenders can defend chances off the rush.
With all of that in mind, while good rush defenders prevent carry-ins by forcing dump-ins, how valuable was the forced dump-in if the other team just recovers the puck back? The most effective defencemen at defending the transition will prevent controlled entries by forcing dump-ins, but they should be able to efficiently recover those dump-ins as well.
Additionally, I track defensive zone puck retrievals off an opposition dump-in, and I split them into three categories; successful, attempted, and failed.
Successful Dump-in Retrievals are defined as successfully retrieving a puck following an opposition dump-in. A retrieval must end in a successful pass to a teammate and/or a successful zone exit to be assessed as successful.
Attempted Dump-In Retrievals, which I may also refer to as opposition retrievals, are defined as plays where the first player going into the corner to retrieve a dump-in is beaten by an opposing player. For example, if a defender and an opposing forechecker are in a race for a dump-in, and the forechecker wins the race and recovers the puck, it’s assessed as a successful dump-in recovery for the opposition, while the defender is assigned an attempted retrieval.
Failed Dump-In Retrievals are defined as plays where the player did win the puck race and retrieved the puck, but following the retrieval, they turn it over.
Defensive Zone
My DZ tracking is a bit more specific.
DZ Puck Touches are the sum of every single puck touch in the DZ. Puck touches are only tracked when a player has evident possession of the puck. Similar to the zone exit criteria, puck touches are only assessed if there was direct forecheck pressure and/or 2+ opposition players were in the defensive zone.
Again, I’d like my tracking to be as reflective of a player’s skill and ability level as much as possible, which is why I place a lot of emphasis on tracking plays under pressure. It’s a lot more difficult to make a successful play while evading a forechecker.
A defensive-zone puck touch can essentially lead to five different outcomes, each of which is tracked:
A controlled zone exit
An uncontrolled zone exit
A pass to a teammate within the zone
A turnover
An icing
Puck touches that don’t lead to any outcome (e.g. a stoppage in play, a penalty, the period comes to an end, etc.) are typically not tracked.
DZ Turnovers are pretty self-explanatory. They’re defined as plays where a player gives up puck possession to the opposing team in the defensive zone.
Alongside specifically tracking retrievals off a dump-in, I will track all DZ Retrievals. This would include dump-in retrievals and other loose puck retrievals; an example of a non-dump-in retrieval would be picking up a rebound off an opposition shot.
Additionally, I also track DZ Breakups, which are defined as instances when a player disrupts the opposing team’s possession, and the player breaking up the play helps the team regain possession. If a player only briefly breaks up a play, while the opposition continues to maintain possession, a breakup will not be assessed.
Final Thoughts
To conclude, these are essentially all the statistics that we will track for the Oilers this upcoming season.
To ensure the accuracy and validity of this project and its data, the timestamp of each event tracked will be noted down.
I plan to frequently use these metrics throughout my articles. Note that I’ve permanently stopped writing regular articles on Substack for quite a while; instead, I’ve been writing weekly pieces on OilersNation for the past two years, which you can check out right here.
If you have any other questions about this project, or any suggestions about other stats that we could potentially track, feel free to let me know! :)
You can find me on Twitter (@NHL_Sid)