I’ve changed the way I’ve watched hockey over the years. As a kid and through my teenage years, I watched hockey like you’d expect any kid to watch hockey: follow the puck and get excited when a goal is scored. I didn’t play a ton of hockey growing up, my parents told me it was too expensive. Therefore some of my ideas about hockey or idioms I follow weren’t ingrained in me by some youth coach at the Pee-Wee level. My interest level in hockey has increased since I’ve gotten older, while my own children became involved in the game.
Real Hockey Men
Since I didn’t grow up in the shadow of a Real Hockey Man (TM), some of the things I heard while learning the game as a fan didn’t make much sense to me. Think of a phrase such as “get pucks in deep”. Well, I guess I understand if you’re trying to be real conservative, which intuitively is what a lot of Real Hockey Men are.
Hockey has historically been a conservative game. Creativity has been seen as a negative; players have been drilled to “make the simple play.” Anchoring bias is a real thing, and no matter how many smart play-driving plays which to the untrained eye aren’t seen as impactful a player makes, one bad turnover that leads to a goal against is what everyone remembers. Hence the desire by coaches for their teams to play conservatively.
Echoing this sentiment last fall, one of the greatest puck wizards to play the game, Pavel Datsyuk, bemoaned the lack of creativity in the NHL. Seriously would you rather watch Datsyuk dangle or Trevor Lewis dump the puck in?
Recall the reputation of P.K. Subban among some of those same Real Hockey Men. The scouting report on him was that he was too risky with the puck, made too many mistakes and turned the puck over too much. Newsflash, guys who have the puck a lot will turn the puck over more than people who don’t have the puck a lot. Witnessing a daily dose of Subban last season shows he likes to have the puck on his stick. Subban has suffered some of the same criticisms that have been thrown at Erik Karlsson; without Karlsson, the Ottawa Senators would be pretty damn bad.
Three Zone Play
Nonetheless, through a lot of manual hard work and some math skills, the hockey community has been examining further the idea of three zone play. Stepping beyond simple possession numbers like Corsi or Fenwick and into what the drivers of puck possession actually are. The seminal work in this field has been done by, who else, Eric Tulsky. Several subsequent feats of math and patience, especially by the aforementioned Tulsky, Corey Sznajder, Dimitri Filipovic and others, have proven the following:
- Controlled Zone Entries generate about twice as many unblocked shots as uncontrolled zone entries
- Preventing Zone Entries is a repeatable skill
- Defenders who generate a Zone Exit with control can move the possession needle
Dividing the ice into three zones forces us to look at how teams gain the offensive zone or exit their own zone. For a team to succeed, ideally they’d have some great puck carriers on their team who can gain the zone with control, and on the other end have defenders who can deny controlled zone entries. Transition play was summed up nicely previously by Dom Luszczyszyn with an analysis of the work done on Corey’s tracking project. I wrote about some of this a few weeks ago when analyzing the impact of the Ryan Ellis injury.
Analyzing zone exits is still an area where the hockey community can improve. The impact on goals can be rather fleeting, because we don’t know exactly what a team is doing in the neutral zone to gain the oppositions blue line. We do know however that players who can exit the zone with control at the very least are giving their team the opportunity to transition the puck from defense to offense. While guys who just dump the puck out to center ice are often on the ice for more shots against because the opposing team simply regroups and brings the puck back into the attacking zone trapping the defenders in the d-zone.
Naturally, then, defenders who can either break up the play at their own blue line, or force the opponent to dump the puck in are far less prone to giving up shots on net. This is another area of analysis that is interesting, yet inconclusive. After all, simply because a defender broke up the play entering the zone doesn’t mean that they’ve actually gained possession of the puck to produce offense; only that they’ve fulfilled the defense part of their job title.
Individual Contributions & Symbiosis
When Shea Weber and Roman Josi were patrolling the blue line together in Nashville, they had a symbiotic relationship. Weber playing the role of stout defender who will battle hard in the corners and clear the front of the net, Josi was the puck mover who transitioned the play up the ice. When two blue liners play so much of the season together, it’s often hard to tease out individual results. We’re usually left with maybe a few hundred minutes where they played separately on the entire season, which presents a sample size issue.
If you pay attention to the rest of the league, you likely heard the comments from Maple Leafs head coach Mike Babcock during training camp. The hockey mad media in Toronto asked him one too many times about keeping winger Zach Hyman with Auston Matthews. He had the following to say:
Well, yeah that makes total sense.
As fans we might want our team to load up one line or one d-pairing with the guys we like watching play the most, but can those guys actually play well together? I’m not going to get into it in-depth here, because it’s outside the subject matter, but here are the defensive-pairings that correspond to the article I wrote about the JOFA line and Fiala.
Presuming we can classify the Predators defensemen into the above category, Alexei Emelin, who will have some big shoes to fill at least until late January, can firmly be classified as a Defensive-Oriented player. The only situation in which such a player can reasonable expect a 50% or more expected goals percentage is in playing with an All-Around defenseman. All-Around D-Men are more valuable when paired with another defensive type, as their expected goals rate is not the highest on the chart if you pair two of them together. We’re looking for symbiosis.
Evaluating the Blue Line
We’ve established that it’s important for a defender to be able to limit zone entries against and to provide controlled zone exits for their team. Ideally, the Predators would have a team full of players who can easily control both entries into and exits from their defensive zone, pushing the forwards up the ice to generate offense on the other end. The salary cap tends to strain under such players.
Using the data from Corey Sznajder from last season, I plotted Nashville’s defensemen and included Emelin’s numbers from Montreal. I divided the results into four quadrants.
- Open Blue Line - Above average controlled entries against and controlled exits for (Fun)
- Closed Blue Line - Below average controlled entries against and controlled exits for (Boring)
- Neutral Zone Liability - Above average controlled entries against and below average controlled exits for (Bad)
- Neutral Zone Play Driver - Below average controlled entries against, and above average controlled exits for (Good)
A few things stand out here. Let’s start with the bad news.
In the Neutral Zone Liability category, we have two players, Matt Irwin and Mattias Ekholm. Having #14 here doesn’t make sense does it? He might be the one result here that doesn’t match with what our mind tells us. Using PuckIQ, Ekholm’s shot attempts allowed per sixty minutes was one of the better marks in the league at 49.7, and 16th best in the league among defenders with 250+ minutes against elite competition. We’d expect those two numbers to trend in the same direction. If Ekholm is a sieve at the blue line how is he so good at preventing shot attempts against? Perhaps its a sample size issue. Corey’s data only has 24 regular season games tracked from last year.
I don’t want to live in a world where the league plays 2-1 hockey every night. Sure it can be exciting especially in the playoffs and I’m not against defense, but scoring chances get fans out of the seats and creates excitement. Moving over to the Closed Blue Line category, among more regular players we have Alexei Emelin and Yannick Weber. These results pass the eye test. Both guys are seen as defensive oriented guys. They can protect their own blue line (Emelin more than Weber) but don’t do much to help the team move the puck up the ice.
Gimme some more Roman Josi! Like Pekka Rinne, the consensus on Josi is polarized. Not that he’s bad, but Tyler Dellow wrote for The Athletic last spring why P.K. Subban is better than Roman Josi. With Josi you have to take some of the good with the bad. In a similar vein to Subban, a guy like Josi who carries the puck a lot and looks to create offense on his own is often viewed in a negative light because he isn’t out there to make the simple play or defend the net front. He’s out there to provide. The eye test matches here as well, with both Josi and Ryan Ellis planted in the Open Blue Line category. The Corsi numbers here line up, both defenders were on the ice for over 115 shot attempts for per sixty minutes.
Then there’s P.K. Subban. He’s the only defender on the team that is above average at preventing entries and above average at generating zone exits. That means he’s really good.
Last Season’s Playoffs
The only confounding narrative with the regular season numbers is Mattias Ekholm. Since it’s a 24 game sample, I’m willing to chalk it up to a sample size issue. Dimitri Filipovic tracked all entry/exit data during the playoffs. This will give us a little bit more information to look at, when teams are putting it all on the line each night. Small sample sizes here, but the Preds played 22 playoff games, almost the exact same amount of games tracked from the regular season.
The numbers here match our perceptions. Sean Tierney made the chart below, and his quadrants are the same as mine, but in different spots. Here Subban, Ellis and Ekholm are all listed in the Neutral Zone Play Driver quadrant, and that’s where we want them to be. Josi is still in the fun category with open play on both ends. Irwin and Weber both ended the playoffs in the Neutral Zone liability category, with none of the Preds defenders ending in the Closed Blue Line category. Intuitively, this reflects well on what we witnessed during their magical playoff run.
If analytics never told us anything that we didn’t already know, they’d be pointless. If they never matched the eye-test, they’d probably be wrong. The sweet-spot is exactly where they reflect some of the things we can see with our own eyes, but they also show us things we missed or that we were never aware of to begin with. With a concept like controlling the blue line, the NHL doesn’t provide anything for the public to generate that data from. We’re left to track those on-ice events on our own. I consulted some of the folks who’ve done some tracking to get an idea of how they do it, in the hopes that my results can be of equal comparison to theirs.
Tracking Zone Exits
- Touches - This is simply the number of exit attempts the defender has. This doesn’t mean how man times the defender touched the puck in the d-zone.
- Exits - Of those touches, how many times does the defender successfully get the puck out of the zone. If the player ices the puck or the clearing attempt leads to a direct re-entry by the opponent, those would both be failed exit attempts. As would a turnover in the d-zone on an attempted exit.
- Possession Exits - Of the total exits, how many times did the team retain possession of the puck on the exit attempt? If the defender carries the puck out of the zone or completes a successful pass out of the zone, those are both counted as successful exits with possession.
- Possession Exit% - Divide the number of successful exits with possession by the number of touches. The higher the number the better.
Tracking Zone Entries
What I’ll be looking at here is how well Nashville defends their own blue line. I’ll be tracking controlled and uncontrolled entries against. Let’s define what each of these things are.
- Targets - How many times does the opponent target a specific defender? The offensive team has to try to enter the zone against a specific defender.
- Carry In Against - Does the opposing team gain the blue line with control of the puck against a defender who’s in a position to influence the play?
- Dump In - Is the opponent forced to dump the puck in or try to chip and chase off the wall? The offense must voluntarily give up possession in order to cross the blue line. Dump-In’s from just inside the red line will not count as a targeted entry. The player dumping the puck in the zone didn’t attack a specific defender and that defender didn’t have to make a play to force the dump-in.
- Break Up - Did the defender break up the entry through a deliberate action? To qualify as a break up the defender has to either with his body or his stick dispossess the offensive player of the puck; which would lead to a recovery by the defense or a forced regroup due to a failed entry attempt by the offense.
- Carry In% - Divide the number of successful carry ins by the number of targets. The lower the number the better.
- Break Up% - Divide the number of successful break ups by the number of targets. The higher the number the better.
I am going to make every effort to complete all Nashville games and stay as current as possible. I’ll post the results on twitter (@LedgerSko) and keep an accumulated total throughout the season. I’ll also be analyzing neutral zone play with this data. Hopefully we can all learn something new about the Predators, their defensive pairings and how exactly the team drives possession in the neutral zone.