All Three Zones Overview: Passing

On Monday, we went over the neutral zone portion of what I’m going to be tracking for the upcoming season. Today, we’re going to move onto passing and shot contribution data. This is all based off a project started by Ryan Stimson, who started tracking passes a couple seasons ago for the New Jersey Devils. He started this project by tracking all passes but has since consolidated it and now only tracks passes which lead to a shot attempt, which makes sense. It has already been established that outshooting your opponent is key to winning in hockey and there is a lot of value in looking into which players are generating those shots.

Shot attempts and on-ice stats like Corsi do a good job of showing who is individually creating shots and how the team is doing with certain players on the ice, but looking at passes & which players are contributing to meaningful possession gives us a closer look at who might be driving the play on a certain line. There is a lot more we can do with data like this too, as Ryan has demonstrated in a series of post on Hockey Graphs. With passing data, we can get a better idea of which areas of the ice teams are more likely to attack, how shooting percentage is impacted on shots with passes, which types of passes lead to higher quality shots, chemistry between linemateshow teams move the puck up the ice on breakouts and how each team defends their own zone. This is just scratching the surface because a full season of this data has yet to be tracked and there’s a lot more questions we could answer by having more data available.

I am hoping to help out with this, as I’m going to be teaming up with Ryan for the upcoming year and will be using his spreadsheets & methods to track passes for this project. Meanwhile, Ryan and his volunteers will continue to track last season so when all is said and done, we’ll have two full seasons of passing data to work with. I will be posting the data here on a game-by-game basis and updating it for every team as the year goes on.

Why is it important and how can you, as a fan, use the data, though?

Why are passing stats important?

I linked to some of Ryan’s work earlier and it does a pretty good job of going over all of the different things we can get a closer look at from this data. The most important thing, however, is that passing stats (or shot contribution as it’s referred to sometimes) does a better job of predicting future goals than just shot generation stats alone. It’s also nice to provide some contexts to the on-ice numbers because it’s easy to say that a player had a bad game because he his team owned only 30% of the shot attempts while he was on the ice, but what went into making that happen? Zone entries help and passing stats only add more context and will give us a better idea of who is contributing the most and who might be a passenger on their line.

What is being tracked?

Tracking passes is a little more complex than tracking zone entries because there’s more codes to remember, but it’s pretty simple once you get the hang of it. What you do is track every time there’s a shot attempt in a game (even ones that aren’t logged in the play-by-play sheet), the time & period where it happened, whether or not it got on goal, the type of shot (wrist, slap shot, one-timer), the last three passes that occurred before the shot, the area of the ice where the pass happened, whether or not it was a scoring chance or if it came off an odd-man situation and if the shot produced a rebound. The score of the game and the goaltender which faced the shot are also tracked. In the end, you’re left with a lot of data that’s a little intimidating to sift through at first.

Here’s what it all looks like broken into different categories.

  • Shot Types: The type of shot that was taken. Categories speak for themselves here (wrist shot, slap shot, one-timer, back-hander, tip/deflection, wrap around or rebound.
  • Passers: Since the last three passes are recorded, passers are broken down into primary, secondary and tertiary categories. We usually only look at the primary passer for individuals, since that is the most accurate judge of talent, but the other categories are looked at for team-wide play, especially on breakouts.
  • Pass Types: For this we record the zone of the ice where the pass began (offensive, neutral or defensive) and the lane where it originated from (left, center & right). These are your basic categories and do a pretty good job of telling you where teams attack. However, Ryan has taken it a step further and included special codes for certain passing plays. Stretch passes and faceoff wins are two of them and they speak for themselves. Passes back to the point are noted to indicate “low-to-high” plays in the offensive zone and passes from behind the end line are used to signify “behind the net” plays. The final category are passes that cross the “Royal Road”, which is an imaginary line that extends from the top of the crease to the top of the faceoff circles in the middle of the ice. Shots with passes that cross this line are typically hard for a goaltender to stop and lead to higher quality chances. They are labeled as “Slot Passes” in my tracking because I hate the term Royal Road and refuse to use it. That’s just a minor quirk from me, though.
  • Rebounds/Second Chances: These are only tracked if the rebound occurs within the home plate area and leads to a scoring chance. They are not logged otherwise.
  • Screened Shots: This is something that wasn’t tracked initially but I’m going to be adding this for the upcoming season because it was suggested a few times and wouldn’t be too difficult to add since I’m tracking almost everything else. It would also be nice to know how much impact this has on a goaltender’s save percentage since it’s something that intuitively makes sense.

Now that we have all of our bases covered, let’s get into what this data looks like at the end and how it can be used to breakdown a game.

Game Example                                                      

Here is what the passing stats look like from the Arizona/Minnesota game I tracked.

pass-1

These are our shot types, which give us an idea of who had the edge in possession and scoring chances. You can see that it wasn’t an eventful game but the Wild had a slight edge at even strength in pretty much every category, especially at generating shots off passes. These stats are boring, though so let’s get into the meat of the passing data:

wid-vs-arizona

Again, Minnesota had their way with the Coyotes when it came to moving the puck around the ice and they had a pretty decisive advantage at getting the puck to the middle of the ice and creating offense from their own zone (stretch passes). The only category Arizona led in was in shots off low-to-high passing plays, which lead to lower quality chances and they couldn’t get much of anything going both in the middle of the ice and in the home-plate area. In other words, the Coyotes pretty much had nothing going offensively. Minnesota wasn’t exactly creating a lot of chances either, but they were much better at creating shots off passes and creating offense in transition.

What led to that? If you look back at my post on zone entries, you’ll remember that I said the Coyotes defensemen struggled to move the puck up the ice and could only generate controlled entries off turnovers or mistakes by the Wild. Conversely, Minnesota had a much easier time rushing the puck up the ice earlier in this game and they seemed to make a point at attacking this way early on.

zucker-1

This is only one minute into the game, so you can get a good idea of what each team’s plan was coming into it. Arizona comes out with a 2-1-2 forecheck while the Wild break the puck out of their own zone. Charlie Coyle currently has the puck and was able to get past the two Arizona forecheckers pretty easily. The Wild now have a 3-on-3 going up the ice and it’s pretty easy to see what they’re trying to do here. Arizona overloaded the right side of the ice, so the forward at the bottom of the picture (Jason Zucker) is going to have a one-on-one situation entering the zone if Coyle can get him the puck.

zucker-2zucker-3

Coyle makes a stretch pass from the neutral zone to Zucker and the Wild get an easy entry. Zucker ends up wasting this by taking a quick shot on goal from where he is standing here, but this gives you a good idea of what the Wild’s attack looked like. Arizona had two forwards in, so there was room in the neutral zone for Minnesota to exploit. They wanted to come up the ice with speed and move the puck laterally to force the Coyotes blue line to defend on the rush. They don’t have the most mobile blue line in the world, so they were going to back off if the Wild moved the puck well enough. Eventually, this is what happened.

dumba-1

Later in the period, the Wild find themselves in a similar situation as they prepare to start a breakout. Things are a little easier this time as both teams are making a line change and there’s only one forechecker in the zone. You can see where the defenseman wants to go with the puck.

dumba-2

Matt Dumba hits Mikael Granlund with a stretch pass, getting the Arizona defense to back off and the Wild are in a good situation here. They have three guys crossing the line with Zach Parise driving the middle lane and Jason Pominville on the weak side as a passing option. They also caught Arizona defenseman Nick Grossmann on his wrong side, so he’s basically a sitting duck and has to concede the blue line.

dumba-3

Parise continues to drive the middle lane and draws in the defense, which opens up a passing lane for Granlund to Pominville. Grossmann didn’t do much to prevent this and the other defenseman was coming off the bench, so he couldn’t do much to prevent it either.

dumba-4

The far-side defenseman finally picks up the play, but it’s too late as Pominville already got a shot off and Parise has positioning on Boyd Gordon in front of the net. Pominville might have been able to hit Parise with a pass, because it’s a tap-in goal if that happens. This isn’t the worst outcome for the Wild, though. They were able to move the puck laterally in the offensive zone, get Anders Lindback moving around and at worst, Parise is in good position for a rebound if he makes the save.

dumba-5

Lindback has to make a pretty difficult save and Gordon does a good job to prevent the rebound. Still a good scoring chance for the Wild here and it was all generated off the rush. This was the start of a good chunk of the Wild’s offense. 11 of their 29 shots started with a stretch pass in the neutral or defensive zones, including their only goal of the game. It gave them a pretty big advantage for most of the game, as Arizona couldn’t generate anything in transition and the Wild are an effective team when they could carry the puck in.

However, if you also look back on the zone entry post, you’ll see that Minnesota only carried the puck in 44% of the time. They were better than Arizona, but that’s still not a very good carry-in rate at all. They also only generated only one shot off a stretch pass in the third period, so Arizona definitely caught onto this and adjusted accordingly. What did they do, though? It was actually pretty simple.

break-out-1

We fast-forward to the third period after Arizona tied the game. Here you see the Wild trying to start their normal breakout and what’s the first thing you notice? All of the Coyotes’ players are in the neutral zone and they have a defenseman (Murphy) playing up on the ice. It’s probably easier for them to do this because they aren’t playing from behind anymore, but the Wild are going to have a tougher time getting through the neutral zone regardless.

break-out-2

Tikhonov shadows Niederreter as the Minnesota forward carries the puck up the ice and has his stick in the passing lane, making it more difficult for the Wild to move the puck laterally like they want to do. Even if he does complete the pass, there’s two players playing up in the neutral zone so Arizona has another road-block set and it’s not going to be a one-on-one situation this time.

breakout-3

Niederreiter ended up making the pass to Koivu but he had an Arizona forward on him the second he received it and the play resulted in a harmless dump-in. Arizona killed the Wild’s transitional play in the third period with this strategy and ended up getting a point out of this game from it. All it took was a slight adjustment in their forecheck and having more than one forward in the neutral zone.

On-ice Data

While the team-wide data is interesting we can take it a step forward by looking at on-ice data for individual players. We saw earlier that Arizona got lit up when it came to giving up passes in the center lane, but who was most responsible for allowing those plays? This helps us get a better idea of the answer.

Passing Lane Corsi

pass-2

PF = Passing Plays For, PA = Passing Plays Against

Pretty ugly night for Arizona in all regards, especially their top line. The most surprising thing here is how much their top pairing of Oliver Ekman-Larsson and Michael Stone were giving up in all three lanes. Their differential in center ice looks the worst because of a lack of offense, but this duo had trouble containing the Wild all night from the looks of things. Ekman-Larsson is thought pretty highly of, so I looked back at the plays to see if there was anything he was doing that led to this. When I did, I noticed that some of it related to their transitional play and how much defensive responsibility is expected from their forwards.

Let’s start with the transitional play:

oel-1

The Coyotes are on the attack now, as Mikkel Boedker just entered the zone with control and pulled up while the rest of his team got onside. He has Ekman Larsson open as a passing option, but Boedker isn’t looking at him and is trying to hit Vermette with a pass instead. It’s not a bad idea, as he could probably set him up for a pretty good shot in that area. The problem is that Boedker isn’t a good enough passer to make this play and there’s a defenseman right in the lane.

oel-2

The pass predictably gets broken up and Ekman Larsson pinched, so Boedker has to hustle to get back and cover up for him. It’s probably a good idea by him because the Wild outnumber the Coyotes in terms of people near the puck.

oel-3

Ekman-Larsson is stuck on the wall while Boedker skates past the Wild player making the pass to start a breakout, so Minnesota is likely going to have a one-on-one situation while entering the zone at worst.

suter-2

This is exactly what happens as Granlund easily gets through the neutral zone and has plenty of room to work with against Michael Stone.

suter-3

Things quickly go from bad to worse, as it becomes a 2-on-1 because Stone tried to play the body on Granlund and he was able to get the pass to Parise in time. Arizona has a forward back, but Boedker is on an island and the other back-checker was too slow to break it up.

suter-4

Boedker doesn’t really do much of anything to break-up the pass and Lindback has to make a pretty difficult save on Pominville. It all started because of a bad pass in the offensive zone and Boedker didn’t exactly do the rest of his team any favors with how he played defensively here. He overskated while covering the point and did a pretty poor job of breaking up the pass to boot, although he is a forward so I guess I should give him some leeway here. Ekman-Larsson’s read wasn’t great here, but Boedker started the whole sequence with a pretty low-percentage play in the offensive zone.

I mentioned that Arizona commands a lot from their forwards in terms of defensive play. We saw this in how they defended the Wild’s breakouts, but it’s a key part of their defensive zone coverage, as well. They typically have a defenseman covering the front of the net while a forward puts pressure on the puck carrier, especially behind the net. A typical situation for them looks like this:

coverage-ari

You can see they have the slot well covered an pressure is being applied to the puck handler behind the net, so there’s nowhere for him to go and he has to make a quick decision with the puck. They left the points wide open, but the chance of a shot going in from there is lower than one from the slot, so it’s a gamble the Coyotes are willing to take.

Every once and awhile a breakdown happens and the Coyotes top line had a big problem with it all night.

granlund-1

The Wild have worked the puck behind the net after retrieving a dump-in. I’ve highlighted two of Arizona’s forwards here, as you’d figure they would either go to the slot or pressure the Wild forward behind the net while Ekman-Larsson patrols the crease.

granlund-2

Things take an interesting turn, as the near-side forward (Vermette) takes Pominville in the slot, Ekman-Larsson goes down to defend the pass and the other forward (Boedker) is just kind of standing there defending nobody. Granlund probably isn’t going to work the play to that side and the other defenseman, Stone, is covering the backdoor. The big change from the last picture is that there’s no one pressuring Granlund behind the net, so he has more time to let something develop, unlike the last picture.

granlund-3

Things break down for the Coyotes here, as Granlund makes the pass to Pominville despite Ekman-Larsson’s defense, Vermette goes out to defend Dumba for some reason, Boedker drifts around and misses Pominville and the forward on the other side beats Stone to the net. Pominville ended up hitting the crossbar here, so Arizona escaped without much harm (and died the game in the next shift), but this could have easily been 2-0 for Minnesota thanks to some spotty coverage and bad miscommunication by the Coyotes forwards. All three of them had pretty bad games in terms of possession and this gives us some good insight as to why.

Individual Stats

Next we’re going to look at individual stats, which is what I’ve done the most work with in my playoff recaps. These are pretty basic and don’t require much explanation, but it is nice to go beyond shot attempts to see who is creating the most offense.

Wild Shots & Passes

pass-3

Breaking down the Wild’s offense like this is pretty interesting because it shows that they were one-line team this game. Parise, Granlund & Pominville carried the load and there’s a major drop off after them. You can also see that Parise & Granlund were doing most of the work on this line too, as Pominville acted as more of a complementary piece. Jared Spurgeon was also a big part of their offense that night, scoring their only goal and having three primary shot assists. With how big stretch passes were for Minnesota, you’d expect their defense to have more shot assists, but Spurgeon, Suter & Brodin seemed to be doing the most here.

I usually include the shot & pass rates, Shot Contribution Percentages and how many times a player setup a certain teammate for a shot, but those are more useful over a larger sample of games, so I won’t include them here. What I can do, however, is break down each player’s shot assists by zone to show how exactly they were setting plays up. This is something that would also be more useful over a larger sample, but it gives us some interesting stuff to work with for game recaps.

pass-4

Parise & Granlund were the Wild’s most dynamic passers this game and Parise did a lot to contribute in terms of transitional play, as well. Pominville also shows up as more of a standout because although he didn’t have many shot assists, he had two stretch passes and that was very important for the Wild. Still, Parise and Granlund did the most in terms of making plays in the offensive zone and in dangerous scoring areas. Not that surprising when looking at their overall numbers compared to the rest of the team.

Again, this will be more useful with more games tracked but I just wanted to give everyone a glimpse of everything.

Conclusion

The passing data is very interesting to work with and having two season’s worth of data will be very important for the hockey stats community. We already know the value of data like this with the little we have publicly available, so there’s no telling what more we can do with it once we have two full seasons tracked. We can build on the discoveries we already made and have a full database and use that to answer new questions that come up everyday. If you’re interested in this type of data or learning more about hockey, then I encourage you to make a pledge to my Patreon or GoFundMe page so I can focus on tracking this coming season and support Ryan’s work with the Passing Project for last season. It takes a crazy amount of work to get these games tracked and the end result is definitely worth the process.

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s