By: Spencer Fascetta Twitter: @pucknerdhockey
A lot has been made about the Bruins’ clear push to embrace a youth movement this season. They actually lead the league in games played by rookies. But, apart from Charlie McAvoy, who has more than lived up to the hype, the B’s 3 other primary rookies (Matt Grzelcyk has been good, but has not played enough to qualify here), Jake DeBrusk, Anders Bjork, and Danton Heinen, really lack the name recognition of the more high profile rookie leaders this year. This is your Matt Barzals, Nico Hischiers, Will Butchers, etcetera. Because of how important young talent has become in this league, and because these three haven’t generated as much hype thus far, I thought I’d look into what kind of value they actually bring to the B’s.
To begin with, I took the data collected through Manny Elk’s Corsica.hockey database. Using NHL.com’s stats page (the only time I will EVER use that abhorrent piece of garbage) to determine the top scoring rookies, I pared my sample frame to the top 25 scoring rookies in the NHL this season as of the winter holiday roster freeze. They are:
Adrian Kempe (LAK)
Alex DeBrincat (CHI)
Alex Kerfoot (COL)
Alex Tuch (VGK)
Anders Bjork (BOS)
Brock Boeser (VAN)
Charles Hudon (MTL)
Charlie McAvoy (BOS)
Christian Fischer (ARZ)
Clayton Keller (ARZ)
Danton Heinen (BOS)
JT Compher (COL)
Jake DeBrusk (BOS)
Jakub Vrana (WSH)
Jesper Bratt (NJD)
Joshua Ho-Sang (NYI)
Kyle Connor (WPG)
Mark Jankowski (CGY)
Martin Frk (DET)
Mathew Barzal (NYI)
Mikhail Sergachev (TBL)
Nico Hischier (NJD)
Pierre-Luc Dubois (CBJ)
Will Butcher (NJD)
Yanni Gourde (TBL)
This includes all 4 rookies of note for the Bruins, as well as a large enough sample size to properly analyze their relative impact across the league. Corsica’s database offers 5 different datasets for individual players: Summary, Relative, Individual, On-Ice, and Context. I took all 5 datasets and combined them into one large dataset to analyze the largest number of statistics possible. Then, using some of my own magic, I cooked up a few graphics to help demonstrate why the quartet of DeBrusk, Bjork, McAvoy, and Heinen deserve much more respect than they are currently getting.
This first graphic displays the individual point distribution per 60 minutes of play. The idea behind a “per 60 minutes” statistic is to account for the effect of playing time on production. A top line player is likely to score more than a 4th liner solely based upon them spending more time on the ice than the 4th liner. Points per 60 and similar statistics show how productive a player is the ice time they are being given. The graphic shows each player’s goals per 60 (red), primary assists per 60 (orange), and secondary assists per 60 (green) stats. Obviously, Brock Boeser is a goal scoring machine. Adrian Kempe appears to be shredding opponents in slightly less ice time than his peers.
As for the Bruins’ 4, Bjork (5th from the top) seems to be relatively well balanced. He looks less productive simply because he is listed directly above the absurdly productive Boeser. However, he is above the median regarding points per 60 production.
McAvoy (8th from the top) appears at first glance to have a really low level of production. This is, however, more of an inditement of his ice time, which will be touched on later. In short, McAvoy is playing significantly more minutes per game than anyone else on this list. So, let’s discount him on this graphic.
Heinen (11th from the top) is, based on this metric, an incredibly productive player. What’s important to note about him on this graphic is how much of his production are primary points. This means he is directly responsible for a majority of the offense he is producing or contributing to, something that is much more likely to be a repeatable action than a gluttony of secondary assists.
DeBrusk (13th from the top) is 3rd (yes, THIRD) in goals per 60 in this group, behind only Boeser and Kempe. He is also 5th in overall points production, behind only Kempe, Boeser, Barzal, and the much, much older Yanni Gourde. For anyone who didn’t think he was worth that 1st Round pick in 2015, consider that he has been a healthy scratch twice already this year. Gourde hasn’t, Kempe was a recall a few games into the season, Boeser has missed only a single game, and Barzal hasn’t missed one for the Islanders. DeBrusk hasn’t had the minutes that Kempe, Boeser, or Barzal has. This is incredibly impressive.
Next, I chose to look at how each player’s zone starts were distributed. The percentage of their shifts starting in the offensive zone are shown in red, neutral zone starts in orange, and defensive zone starts in blue. Coaches will often shelter younger or less defensively responsible players by starting more of their shifts in the offensive zone. Bjork is a prime example of this. He is not on the penalty kill in Boston, and no rookie in this group starts a higher percentage of their shifts in the offensive zone. The only one who starts a lower percentage of their shifts in the defensive zone is (curiously enough) Tampa’s Mikhail Sergachev, one of only 3 defensemen on this list.
Similarly, DeBrusk sees a majority of his shifts start in the offensive zone, and starts the fewest number of his shifts in the defensive zone. Where this gets interesting is Heinen. Heinen is arguably the most balanced individual on this list. There is still a slight advantage to the offensive zone, but his zone starts are almost entirely equally distributed. There is tangible evidence for this, as he is the only one of the B’s 3 big rookie forwards who consistently sees time on the penalty kill. It appears he has Bruce Cassidy’s trust in his own end. McAvoy also sees a large number of his shifts start in the offensive zone, and has the same offensive, neutral, defensive progression as Bjork and DeBrusk, but it is much less pronounced than those two, and he starts more of a percentage of his shifts in his own end than the other two defensemen on this list, Will Butcher, and Mikhail Sergachev.
This is quite possibly one of my favorite graphics I made. This is a model of a player’s shot efficiency. It compares their shooting percentage to the shots generated for per 60 minutes of play. Players to the upper right of the graph are producing lots of shots, and finishing at a consumer rate. The bottom right quadrant is full of players who are shooting at a higher than normal shooting percentage and will, in all likelihood, see their percentages regress towards the mean. The opposite is true in the top left quadrant. These players are generating a ton of shots, but just aren’t finishing at the rate at which one would expect them to. The dotted vertical line is the average NHL shooting percentage, 9%. You might be asking, “Why are they all different sized dots?” Excellent question. The dots are all scaled based on the percentage of overall ice time skated by the team that the individual player skated, or TOI%. Players with a large dot are relied on to play more minutes for their team, whereas the tiny dots represent players who regularly see time on the 3rd or 4th line for their respective team. Each of the B’s 4 rookies has been labeled.
Now, what do each of their relative positions on this graph tell us? Well, only DeBrusk is shooting above the league average, at 10.64%. Bjork and Heinen are relatively close to the average of 9%, at 8.7% and 8.81% respectively. McAvoy is a defenseman, so his shooting percentage is likely to be lower than the average. He is still hanging around the group average though, at 7.85%, a very impressive rate for a rookie blueliner. All three forwards are at or above the average shooting percentage for the group. Heinen and Bjork are almost at the exact center of the graph, while DeBrusk is actually producing slightly fewer shots per 60 minutes, but appears to be a slightly better finisher at this juncture. Also, McAvoy has the biggest dot. Just thought I’d throw that out there again.
So, we already looked at shooting efficiency, why not how efficient they are offensively in general? This graphic uses the same size scale (TOI%) and compares a player’s goals for percentage (the goals they factor into for and against their team divided by all goals scored for and against the said team) to their EXPECTED goals for percentage. Players in the bottom right are underproducing according to this graph (i.e., their expected goals for percentage is higher than their current one by a considerable amount), and those in the top left are overproducing their expected rate.
All four Bruins fall to the right of the 50% mark regarding actual goals for percentage. This means all 4 of them help produce more goals for their team than they allow while they are on the ice. I feel like that might be relatively important. McAvoy and Heinen are both well into the “good” quadrant, and Heinen has the highest expected goals for the percentage of any rookie on this list. DeBrusk and Bjork both fall into the “underproducing” category, which tells me they are probably going to start to score more.
Of note, Heinen’s expected goals for percentage is nearly 60%, meaning, based on how he has played, his team should expect to accumulate 60% of the goals scored while he is on the ice. That’s ridiculous. What’s funnier is McAvoy’s actual goals for percentage is OVER 60% (although his expected rate is closer to 55%, still quite good).
Final graphic with this group of rookies. This once again uses the TOI% size scale and compares a player’s Corsi For per 60 to their Corsi Against per 60. In short, how are the shot attempts being distributed while they are on the ice? I will let this one speak for itself a bit because Corsi is relatively self-explanatory. A higher number equates to a higher percentage of the shot rates they control per 60 minutes of ice time. “For” is for their own team, and “against” is against their team. Therefore, the Y-axis on this graph is offensive shot attempts, and the X-axis is defensive shot attempts they allow. The dotted axes are the expected average of 50%. I’d say that DeBrusk falls into the “FUN” category (lots of shot attempts for AND against), whereas the other three are relatively defensively responsible. Bjork has the highest Corsi For per 60 rate of the four at 61.81. Interesting…
Now you should have a pretty good idea of how McAvoy, Bjork, Heinen, and DeBrusk all stack up against the league’s best young guns. But what about in the Black and Gold?
Remember that Offensive Efficiency graphic earlier? Here’s the Bruins’ distribution. I have labeled each of the four we have been discussing, as well as the B’s top line of Marchand, Bergeron, and Pastrnak, all of whom fall into the “ludicrous” category, and Adam McQuaid, just because he is so clearly the worst of the group despite getting a sizeable chunk of the ice time. This time, I compared shots for the percentage to goals for percentage. As you can see, all four rookies fall into the “Good” quadrant and are better than more than half of the B’s roster. Of note is Matt Grzelcyk, who is unlabeled, but represented by the bright pink dot in the far top right. While he’s on the ice, the team is producing 65% of the shots, and around 77% of the total goals. If he had played more, he would also be a part of this analysis. It also demonstrates clearly that he should not come out of this lineup in favor of Adam McQuaid, who is absolutely abysmal and has not been on the ice for a goal for this year. Also, Brad Marchand is producing almost 91% of the total goals scored while he is on the ice. I’m sorry, what????
So, in short, you really ought to take Heinen, Bjork, and DeBrusk seriously. McAvoy has gotten a bit of Calder attention based on the sheer amount of minutes he is being asked to play, and the competition he is being asked to play against, but the other 3 really haven’t been mentioned at all. DeBrusk and Heinen should absolutely be getting legitimate Calder consideration, and Bjork isn’t very far behind. And Don Sweeney hit it out of the park by committing to this youth movement this year. All of these kids appear to indeed be alright. *Insert really ham-fisted Capri-Sun Pun here*
All data courtesy of Corsica.hockey. Data scraped on December 24th, 2017. Master data tables and graphics created by Spencer Fascetta (aka PuckNerd) utilizing Microsoft Excel and Tableau.
Please give me a follow on Twitter (found above) and check out/subscribe to my YouTube Channel (@PuckNerd) for more of my content!