By: Spencer Fascetta Twitter: @pucknerdhockey
More than the “Trade Tuukka” chants, or the “Deal Krug” narrative, I have been most perplexed and frustrated by the trend of Bruins fans to constantly bring down Brandon Carlo. Carlo is a bonafide Top 4 defenseman in this league, does plenty well, and as soon as he makes a 21-year-old mistake (oh yeah, buried the lede here – HE’S STILL ONLY 21), people jump on the bandwagon and scream from atop the highrises in Boston to get rid of him as fast as possible. To me, this is twofold. For one, Boston fans are consistently impatient. They don’t like waiting too long for players to develop and are quick to bash youngsters for their flaws. This is likely due to how spoiled they have been as a fanbase. When you are exposed to Bobby Orr, Ray Bourque, Zdeno Chara, and now the emerging Charlie McAvoy, fair weather fans have a tendency to assume that every young defenseman should be this way, or they are automatically a bust.
No, they are not.
The irony is until McAvoy came along, Carlo was the beloved d-prospect who fans didn’t want to be dealt for anyone short of Connor McDavid, and three first-round picks (still may want to check on that, Pete Chiarelli’s starting to get desperate). As soon as McAvoy broke onto the scene, well, Carlo just wasn’t Chuckie. Newsflash: they have completely different skillsets, and that’s OK.
The second reason I think is that Carlo stands at 6’5″, 203 lbs, and doesn’t pile up highlight-reel, bone-crushing hits. I like to call this “Dougie Hamilton Syndrome,” only, Carlo hasn’t put up the pure offensive numbers that let some fans stomach Hamilton’s perceived lack of physicality. This is a fundamental flaw in how people perceive defense should be played. You do not have to hit people to be a good defenseman. You do not have to hit people because you are big. Carlo’s game is much, much more than that. But, you try explaining that to the fanbase of the Big Bad Bruins. They always have time (and wayyyyy too much money) for the grizzled veteran who will knock an opponent’s face in (I see you, Adam McQuaid).
By this time, you know my schtick – present an argument I find ridiculous and provide plenty of graphical information to support my hypothesis. In short, nerd stuff. Might I say, PUCKNerd Stuff? Yeah, I know, humor’s not really my forte. Anyhoo…
I looked at the last two years of Boston defensemen since Carlo broke into the league and only looked at players who suited up for a minimum of 500 minutes in the black and gold. This limits the dataset to guys who are heavily relied upon and likely played Top 4 minutes for a long stretch in that timeframe. There are only 7 players who qualify: Zdeno Chara, Adam McQuaid, Charlie McAvoy (yeah, ALREADY), Kevan Miller, Colin Miller, Torey Krug, and, yes, Brandon Carlo. Below is a distribution of their zone starts.
Red represents the percentage of shifts started in the offensive zone, orange in the neutral zone, and blue in the defensive zone. You may be interested to note that Carlo has the second lowest percentage of his shifts start in the offensive zone, and the second highest percentage of his shifts start in the defensive zone. The only player getting a higher chunk of his shifts starting in his own zone is Zdeno Chara. Mind you, Carlo broke into the league as a 19-year-old. This says that not only do the Bruins trust him in his own zone, they rely heavily on him to get the job done.
Well, that’s great. But a zone starts distribution that favors defensive zone starts does not necessarily equate to a GOOD defensive player. So, let’s look at Corsi For and Against per Hour. This normalizes Corsi rates for ice time, though the size of the data points equates to the percentage of total time on ice of the team’s time on the ice each player was deployed for.
Only Chara and McAvoy see a more significant percentage of the team’s ice time than Carlo, and he is their best Corsi Against player by a decent amount. Not only that, but he is also a net positive in Corsi, as he falls favorably within the “good” quadrant. What does this indicate? When he is on the ice, the Bruins are producing more shot attempts than the other team, which is indicative of better puck possession. This made me a bit skeptical, as our good friend Adam McQuaid ALSO shows up in that quadrant. So, what if we look at the TYPE of minutes Carlo has been asked to play?
Looking at the Time on Ice percentage adjusted for quality of competition, and comparing that to shot suppression, or Corsi Against per Hour, also adjusted for quality of competition, you start seeing a better picture. Pay close attention to the axes. The Corsi axis (the x-axis on your standard graph) is between 55% and 58% Corsi – an average possession player will fall around 50%. Based on this, Carlo appears to have played the 3rd most against the hardest competition. This makes sense, as it is supported by the previous chart. He is still good at suppressing the opposition, as his Corsi Against per Hour is third lowest of this group – meaning he actually has the third lowest number of shot attempts allowed while is on the ice of this group. Curiously, Colin Miller and Adam McQuaid are the only ones who are better at shot suppression but play significantly easier minutes and a much lower number of minutes. Also, Charlie McAvoy is the definition of fun, but he is playing easily the most difficult opposition of this group as a 21-year-old rookie. That’s objectively ridiculous. Please give him the Calder NOW.
At this point, some of you might be saying, “Carlo is obviously a product of who he plays with.” AHA. I have out-thought you because I already have that information for you.
Since making his NHL debut, Brandon Carlo has played significant minutes (more than a couple of games) with only four players: Torey Krug, Zdeno Chara, John-Michael Liles, and Kevan Miller, with the majority of his time being spent with (by far) Chara and Krug.
The 50% line on this chart represents an even distribution of zone starts. Those falling below the line favor defensive zone starts, those high favor offensive zone starts. As one would expect, Carlo has seen a lot more aggressive zone starts when paired with Torey Krug. With the other three, he was hemmed in his own end quite a bit.
Let’s check out that Corsi For versus Against per Hour graph again, but this time, look at how each pairing performed. The Liles/Carlo pairing was objectively bad, but the other 3 have been reasonably good. Krug/Carlo is far and away the best pairing of this group, and they’ve played enough of a sample size of games together to indicate that this is real. In fact, looking at the PDO values for each pairing, each of which suggests how repeatable their performance is (PDO = save percentage + shooting percentage; expect most PDO values to trend towards 100.0), you see that the three best pairings actually seem to be about where they should be. The Liles pairing, let’s chalk that up to limited sample size.
Now, I checked out how efficient each pairing was regarding offensive production. I compared expected goals differential to their produced goals differential to do so. Trending towards the top right of the graph is good, towards the bottom left is kind of bad, and any other direction indicates that they are not performing as expected.
The Chara/Carlo pairing is clearly the best, buoyed in large part to a high expected goal differential. The Krug/Carlo pairing is actually underperforming, while the other two pairings, well, they weren’t tremendously good.
Great. So, Carlo is a good defenseman – when compared to his own teammates. Despite Boston’s reputation as one of the more defensively stout teams in the league, that doesn’t mean a whole heck of a lot. So, let’s look at Carlo in comparison to all NHL defensemen who have played over 1000 minutes since his NHL debut.
Looking at a comparison of the difficulty of minutes and shot suppression ability per hour, as we did for the partners Carlo has played with, Carlo matches up quite favorably to some of the NHL’s elite. I have pointed out Erik Karlsson and Brent Burns, the two most recent Norris Trophy winners, as well as Marc-Edouard Vlasic, who plays the most difficult minutes of any defenseman in the league by far, Yannick Weber, who has played the easiest minutes of this group by a large margin, as well as Fedor Tyutin, who has gotten his teeth caved in the most of anyone in this dataset. Carlo is firmly in the top 1/4th of the group, indicating that he is, at worst, a #3 defenseman in this league.
Now, Corsi For and Against per Hour. How does he stack up? Well, Burns is ridiculous, Torey Krug is (unsurprisingly) a Top 5 offensive defenseman in the league, Morgan Rielly is quite good, and Carlo is well into the “good” quadrant. I think I will take him on my defense corps.
What does all of this do to the trade Carlo argument? Well, Brandon Carlo is demonstrably a good defensive defenseman in the NHL at the ripe old age of 21. He’s a right-handed defenseman who is 6’5″ and skates incredibly well. He defends quite well in his own end and plays a very cerebral game on the back end. If you trade Carlo, you immediately will be looking for another Brandon Carlo. The good news is, he won’t need to play top pairing minutes in Boston, which allows him to dominate in a slightly lesser role, what with Charlie McAvoy looking very much like perennial Norris Candidate in his own right. So, please. Just because he makes some weird mistakes, let him figure it out and back off. The end result is going to be one you want to stick around for.
All data mined through the databases on Corsica.hockey. Collected as of January 17th, 2018. All graphs are courtesy of PuckNerd and are not to be used without the express written consent of myself. Thank you.
Please give me a follow on Twitter (found above) and check out/subscribe to my YouTube Channel (@PuckNerd) for more of my content!