Not terribly long ago I wrote about who Larsson was not. He’s not Victor Hedman, specifically. Should Larsson become that kind of a player then we should consider ourselves lucky, because he hasn’t shown that career path in his development. What Larsson has shown is really very odd, actually. I think we can expect a shutdown defender, but there’s more to this story that makes any prediction pretty fuzzy.
After a somewhat promising (if not predictably slow) rookie campaign in 2011-2012, the young Devil Defender started his sophomore season in the AHL. That’s no slight to him, it was the lockout. The problem was that after the lockout ended he returned to the NHL and his offense stayed behind. Over the next 63 NHL games he played between the lockout shortened 2012-2013 season and the following 2013-2014 season, Larsson produced just 1 goal and 8 assists. He spent more time in the AHL.
It wasn’t really until 2 seasons ago that he re-established himself as a fulltime NHL option in New Jersey, but when he did he vaulted to the top pairing with Andy Greene. It’s not how long it took him to become an honest-to-goodness NHL player that makes Larsson a difficult player to forecast. It’s actually how he was used in New Jersey once he reached that status.
USAGE
One of the things that not all fancy stats analysts in the public sphere agree upon is how much quality of competition and zone starts matter when it comes to the effects on shot attempt ratios and other metrics. Some have very convincingly argued that the rate at which players see different tiers of competition doesn’t actually change depending on line or pairing (when looked at as a whole). Some suggest it can impact things greatly. Certainly the traditional hockey thinking is that shutdown defenders will see the top pairing more and play in the defensive zone more. As a result there are fewer opportunities to be on the attack.
Frankly, both sides have really valid arguments and my opinion is that for the most-part NHL defenders spend relatively comparable percentages starting in the offensive zone, defensive zone, neutral zone, and quality of competition is probably overstated compared to quality of teammate. For example, Eric Gryba and Andrej Sekera have seen relatively similar ratios of offensive zone starts, defensive zone starts, and neutral zone starts over the last 2 years, but Gryba’s most common linemates (skaters only) have been Mark Borowiecki, Brandon Davidson, Erik Condra, and David Legwand. Sekera’s most common linemates were Justin Faulk, Mark Fayne, Taylor Hall, and Darnell Nurse. Perhaps not a great surprise that Sekera’s fancies are better given that Sekera shows a greater roundedness to his overall game but it helps having good playing partners.
In any event, a lot of NHL defenders have those similar ratios for zone starts and quality of competition, but Adam Larsson, over the last 2 seasons especially, has had extreme deployment from the New Jersey Devils. It’s so extreme that while it’s easy to compare other defenders like apples-to-apples, Larsson’s experience in New Jersey might make it like comparing apples-to-oranges. That’s what makes predictions about what we might expect from him so imperfect.
Over the last 2 seasons Adam Larsson has played 2523:31 5v5 over 146 games. If we look at the list of all NHL defenders who have played at least 2500 minutes over that time we get a list of 57 total defenders. We have the information, we can see the total number of faceoffs a defender has been on the ice for, how many were in each zone, and what percentage of their individual totals each one makes for them. For example, PK Subban has been on the ice for 2995 total faceoffs over the last 2 seasons, 970 of them were in the OZ, 940 of them were in the DZ, his OZFO% is 32.4%, DZFO% is 31.4%, and NZFO% is 36.2%.
LARSSON'S ZONE STARTS
Of those 57 defenders, Adam Larsson is dead last in total offensive zone starts with 526. Number 56 in the list still received more than 100 additional offensive zone starts than Larsson. His primary playing partner on the Devils (Andy Greene) was 55th with 649 offensive zone faceoffs. Players who made up the middle tier of the 57 (so from 20th to 38th) ranged from 898 to 792 offensive zone faceoffs. So Larsson’s lack of offensive zone starts is so extreme that the difference between him and the next closes player is the same as the difference between 19th and 38th. That’s extreme.
If we do the same exercise but for defensive zone starts then Larsson ends up in 12th of 57 for the most number of faceoffs started in his own zone. Over the last 2 seasons Larsson has started in his own end 1005 times. This is basically the same number of times that Doughty, Hjalmarsson, and Ekman-Larsson have started in their own zone. Here’s the difference: All of those other 3 defenders have been on the ice for more than 3000 total faceoffs. Larsson has only been on for 2396 faceoffs.
When we express Larsson’s usage as a percentage the extremity of his usage becomes even more apparent. Adam Larsson had the lowest percentage of offensive zone starts of the 57 defenders who played 2500 minutes or more over the last 2 seasons. The young defender only started 22.0% of the faceoffs he was on the ice for in the attacking zone. The only player in the league with comparable usage was his partner Andy Greene (22.8%).
Again, the difference between Larsson’s OZFO% and the next closest skater (who isn’t his partner) was quite vast. Larsson was at 22% and Marc Staal was 55th with 25.7%. That’s a difference of 3.7% (if you don’t want to do the math). Once again that’s larger than the difference between the entire middle tier of defenders between 19 and 38.
In terms of his defensive zone faceoff percentage, you can probably guess what I’m about to say. He had the highest percentage of DZ starts among all NHL defenders who played more than 2500 5v5 minutes over the last 2 seasons. His 41.9% DZFO% is astronomical. He and Greene were the only two defenders in the league even in the 40’s. The next non-partner on the list was Rasmus Ristolainen with 38.5% DZFO% and that’s a difference of 3.4 percent.
SO WHAT?
What I’m trying to show here is that Larsson’s usage was extreme. It was so extreme, in fact, that I think trying to compare him over this time to anyone is not going to yield very reliable results. It’s true that his fancy stats have not been stellar, but exactly how much of that is the result of his usage? I don’t believe he will be subjected to those kinds of extremes as a member of the Oilers. The lowest percentage of OZ faceoffs by regular defenders last year was Fayne’s 26%. The highest percentage of DZ faceoffs was Nurse’s 34.9%. Klefbom had relatively even distribution of zone starts. I doubt the coaching staff will completely change their deployment as a whole and for Klefbom just because Larsson is paired with him.
I think it’s a good bet that Larsson’s extreme deployment will come to an end this fall, then we might be able to see if he can flourish without the crushing grind of own-zone starts he’s been party to over the last 2 seasons. When I look at his metrics right now, nothing stands out as immensely positive except for his ability to limit shot attempts. If I were to place the bet on anybody that their raw percentages will change drastically next year, it will be on Larsson and it will be based very much on how differently the Oilers will use him next season.
Follow me on twitter @Archaeologuy
Edmonton Oilers Blogger • 
Join the Discussion:
»