Month: December 2019

In Defense of “Stay-At-Home” Defenders (Or: How I Learned to Stop Worrying and Love the Bahl)

In Defense of “Stay-At-Home” Defenders (Or: How I Learned to Stop Worrying and Love the Bahl)

To get to my point I have to spend some time on the backstory.  The New Jersey Devils traded Taylor Hall.  I’ve made no secret that I am a Devils fan.  So, this trade hits very close to home for me.  While pundits and fans alike battled over what the return should be, I dug my heals in that the Devils needed to acquire, as a part of the deal, a really strong “shut down” defender.  When they acquired prospect Kevin Bahl, I was quite happy.  I got an alert about the trade while out at the mall, Christmas shopping with my fiance.  My initial reaction was, “They did it.  They got Bahl.  YES, perfect.”  By the time I got home, I had friends messaging me asking my input and they were shocked at my thrill.  A scan of twitter and the hockey “blogosphere” told me why.  A cascade of buzzwords flooded hockey twitter.  Phrases like “NHLe” and “no-offensive upside” dominated the conversation.  My moment of thrill and joy was shattered.  So if I sound like I took this trade a little personally for whatever reason, this is why.

The definition of a stay-at-home defender really depends who you’re talking to.  To many folks, the concept is outdated, and denotes a defender who isn’t top 4 potential.  That’s only partially true.  The truth of the matter is that there’s a gap in the talent caused by a few factors.  The NHL has changed the way it calls hooking, slashing, and interference in ways that favor the offensive player.  It’s all part of the NHL’s long quest to increase scoring.  These defenders who learned to defend under the old rule sets are having their tools taken away from them, and thus their effectiveness wanes.  Working in tandem with the rule changes is the fact that the speed of the game has increased.  With the game being faster, the big, slow defender has been left in the dust.  We see this around the league.  Scoring is up, but we also see numerous fan bases (The Devils included) clamoring for improved defense.  There just doesn’t seem to be enough talented, defensively capable, stay-at-home defenders available to go around.  The rule changes have hindered the effectiveness of some members of the current generation of NHL defenders as well as the prospects who came up playing their junior hockey under the older rules.  They learned and developed to play a style of game that no longer exists.  This is where your Samuel Morins and your Jamie Oleksiaks (and to some degree Erik Gudbranson) come into play.  They are big (6’6 and 6’7 respectively) defenders designed for the NHL of 2013, not 2020.  Therefore, in many eyes, the concept of the stay-at-home defender is also a concept not made for the NHL of 2020.  I don’t agree.  Instead, the talent level, skating ability and hockey IQ of the next generation of stay-at-home NHL defenders has to elevate to be able to accommodate today’s game.  Kevin Bahl is part of that next generation.

The “stay-at-home” or “shut down” defender also is a victim of modern analytics.  I engaged in a twitter exchange recently with Byron Bader, the creator of the NHLe Player Comparison tool.  Bader is someone whose opinion and work I have an immense amount of respect for and he’s a very good twitter follow.
bader exchange cleaned up
I talked to him about this tweet and he was gracious enough to respond.  In our discussion I asked if the lack of offensive upside would cause a big hit to NHLe rating.  He informed me that it definitely does.  With this information in hand, I decided to use his
NHLe Comparison tool to do some research of my own.  I went for comparisons of Bahl to players whom I would consider “successful shut-down defenders”

defensive defender comparions
You’ll notice under the players’ names the words “average producer” and “replacement producer.”  This brings me to my problem with the NHLe model for rating this kind of player.  According to the NHLe model,  Nikita Zadorov is a replacement level defenseman.  Chara and Johnson are “average.”  Chara is mostly here just for the sake of the outlier comparison although I don’t expect Kevin Bahl to be the next Chara.  However it does not seem unreasonable, given his size and skillset, to expect that he might be able to be the next Zadorov or Johnson.  In fact, in watching film he plays a lot like Nikita Zadorov.  That said, based on the NHLe model, Zadorov is “replacement level.”  If you watch the Colorado Avalanche at all, and you get the opportunity to watch how effective Erik Johnson and Nikita Zadorov are, you will understand where I’m going with this.  Colorado is an incredibly successful team right now and they’re very difficult to play against.  Zadorov and Johnson are a big part of that.  Their contributions to their team are beyond what I would consider “average” or “replacement level.”  There aren’t many like them in the league right now.  They’re big, strong, physical defenders who are also able to skate with forwards and take away offensive chances.   I challenge any hockey fan to watch these compilation videos of Zadorov and try to claim that is not a player they would want on their team.

The issue at hand is that the analytics crowd would tell you that Zadorov is a 3rd pairing defender.  However, his usage and minute consumption is clearly that of a top 4 defender.  Among Avalanche defenders, Zadorov averages the 3rd most minutes at 5 on 5 and the 4th most on the penalty kill.  Erik Johnson is 2nd in 5 on 5 and 1st on the penalty kill.  The actual usage, and contributions of these players, far exceeds their projections based on the NHLe analytic model.  This is because the model is based very heavily on the offensive production and contributions of a player.  I haven’t focused in on Johnson as much as Zadorov here, and that’s because Johnson is the pinnacle of what a stay-at-home/shut-down defender can be in the modern NHL.  Through 12 NHL seasons he has averaged over 22 minutes per game and often playing against the toughest competition the NHL has to offer.  All this while offering, give or take,  30 points per season.

My film review of Kevin Bahl tells me that he’s a player that is closer to a Nikita Zadorov than a Samuel Morin.  This type of player is becoming an endangered species in the NHL.  But they are still players that coaches, teammates and fans want to see, and CONVERSELY, opponents DO NOT want to see.  I’m not attempting to conclusively say that Kevin Bahl is the next Nikita Zadorov.  I am saying that my expectations for Bahl are based around the fact that this is the type of player he could become.  Because they are an endangered species, this type of player is incredibly difficult to find.  Trying to draft one, or project one, is always going to be a gamble.  However, the potential payoff, particularly for the New Jersey Devils, of having a big, strong, hard hitting defender to sure up their defensive zone play in the future, is a gamble that was worth taking.

The New Jersey Devils: An Analytics Take

Look, I get it.  Not everyone is ready to embrace analytics in making their lineup decision.  But, when you start the season like the New Jersey Devils have, and you make a coaching change before game 30, you have officially reached the point of “seeing what sticks.”  There’s going to be no shortage of armchair GM opinions and commentators offering up fixes, so here’s my contribution to that noise.

The Hypothesis: The New Jersey Devil’s forward line with Jack Hughes should be:
LW: Blake Coleman
C: Jack Hughes
RW: Nikita Gusev

The New Jersey Devils are an analytical disaster with the exception of about 2 players.  Nikita Gusev and Blake Coleman are not exactly tearing up the league, but they are the only Devils players with a Game Score per 60 above 2.00 at 5 on 5.  For the majority of this discussion I’m going to focus solely on analytics as they pertain to 5 on 5 play.

(Game Score explanation for those uninitiated)

The Game score ranking of Devils centers (at 5 on 5) looks like this:
Nico Hischier: 1.66
Jack Hughes: 1.41
Travis Zajac: 1.00
Pavel Zacha: 0.55 (sheesh)
Kevin Rooney: -0.19 (WOWBAD)

Where am I going with this?
Travis Zajac should not be centering Coleman and Gusev.  It should be Hughes.

“Well yeah but that’s just Game Score and that’s not even a real stat and you’re a hack and you don’t know what you’re talking about.”

Lets look at some of the other analytics here.
Corsi
A short version of the explanation of Corsi is that it’s a metric that divides the scoring chances for a team while a player is on the ice, by the total number of scoring chances that occurred while that player is on the ice.  50% means a player breaks even.  The Devils are a TERRIBLE possession/corsi team.  To that end there are only 2 players on the roster with (5on5) corsi’s above 50%.  They are Kyle Palmieri and Jack Hughes.

You can dig through all the Devils analytics available on Corsica Hockey and you will find a trend.  In more than a few individual metrics, Jack Hughes is as good or better than Travis Zajac.

But we can’t just run off with our analytics in hand shouting from the rooftops that we have the answer.  What about the eye test?

Watching the Devils,  some of their best scoring opportunities are coming when Coleman and Gusev are on the ice.  Travis Zajac is still a good 2 way center and is very valuable to the Devils in a bottom 6 type role and as a penalty killer.  The issue here becomes, Coleman – Zajac – Gusev are deployed in more of a 3rd line role, when the wingers on this line are 2 of the only people who have been able to produce points for the woeful Devils at 5 on 5.  I imagine that production is a reason that the Devils coaching staff has been unwilling to tamper with this line.  Jack Hughes is offensively talented in ways the Travis Zajac is not.  Jack Hughes is a top tier talent in terms of offensive creativity.  The idea of keeping him with the forwards who have been the Devils best point producers for 3 years (Hall and Palmieri) certainly would seem to make a lot of sense.  You’re putting players with a history of production around a young, talented center in the hopes they can get going with each other.  Also, Hischier has been out with illness and Hughes is the best option with him out.   A lot of Hall and Palmieri’s production comes from the power play.  If you isolate all their scoring to just 5 on 5 game state, an interesting picture emerges.  When Hischier comes back, he should go back to centering Hall and Palmieri, while Hughes is moved to line with Gusev and Coleman.

Over the last 3 seasons, only 8 Devils forwards have played more than 1500 minutes of 5on5 ice time.  These are: Nico Hischier, Taylor Hall, Blake Coleman, Miles Wood, Kyle Palmieri, Travis Zajac, Jesper Bratt and Pavel Zacha.  Remembering that this year has Taylor Hall’s MVP season in the mix, his individual expected goals for per 60 minutes, at 5on5, was 1.01.  Blake Coleman, whom does not have an MVP season, is in second place among these 8 forwards and has an individual expected goals for of 0.9.  For context Nico Hischier is at 0.85 in third.  Coleman’s numbers through many 5 on 5 analytics are in the top 3 or top 4 among these 8 forwards.  The data of the last 3 years says that Blake Coleman is one of the top 3 wingers for the Devils, and he might even be the 2nd best wing the Devils have.

Blake Coleman’s production at 5 on 5 would suggest that he could/should be a top 6 wing on his team.  His production this year has continued to be strong with line mate Nikita Gusev.  Travis Zajac, however, is not a top 2 center on Devils, in terms of 5 on 5 production.  The other side of the coin here, is that Hall and Palmieri have seen their best production with Hischier as their center.  The Hall trade could really shake things up for the Devils, but before they move ahead with that, I think they need to give themselves an idea of who else on this team can produce points for them.  They can start this process by putting a more offensively gifted center (Jack Hughes) with Coleman and Gusev.