A lot has been made this year about P.K. Subban’s sensational season, and whether he’s played the right kind of minutes to be considered a contender for the Norris Trophy.
More than minutes play into whether a player is a candidate for an individual trophy, but what I’d like to do is take a look at the most recent Norris Trophy winner, Erik Karlsson, and see how his Norris winning season compares to Subban’s.
I want to divide the analysis into three separate areas; performance, which will take into account point production and possession; role, which will look at the level of competition each player faces and usage; and each player’s defense partner.
I’ll be using a large variety of statistics in this examination, a glossary of which is available here.
A special thank you to Gabe Desjardins and his site Behind the Net, and David Johnson and his site Hockey Analysis, without which this article would have taken a month to write.
Part 1 – Performance
Subban has been undeniably dominant this year. He leads all defensemen in points and goals, leads the entire league in powerplay points, and is a dominant possession driver. Using a variety statistics it becomes necessary to remove a little noise, so in order to get a better concept of all the stats I’ll be using, I’ll also be including league rank. However to eliminate small sample sizes, any defenseman with under 500 even strength minutes played in 2013 will be eliminated, and the same goes for any defenseman with under 1000 even strength minutes played in 2011-12. This limits each sample to 126 defensemen in 2013, and 121 defensemen in 2011-12.
| League Rank | Subban | Statistic | Karlsson | League Rank |
| 5 | 65.9% | Goals for percentage | 55.2% | 21 |
| 2 | 60.5% | Shots for percentage | 53.2% | 23 |
| 2 | 59.8% | Fenwick for percentage | 53.7% | 19 |
| 3 | 57.5% | Corsi for Percentage | 54.9% | 11 |
| 5 | 16.4 | Relative Corsi | 11.3 | 3 |
I could write up a big long paragraph about the differences between the two players here, but the league ranks speak for themselves. Karlsson is an amazing player, but this season Subban is on a whole other level in terms of possession.
| League Rank | Subban | Statistic | Karlsson | League Rank |
| 1 | 2.59 | Points per 60 | 2.28 | 1 |
| 1 | 0.79 | Goals per 60 | 0.56 | 2 |
| 21 | 1.07 | Even strength points per 60 | 1.72 | 1 |
| 2 | 7.69 | Shots per 60 | 7.64 | 2 |
| N/A | 4.51 | Even strength shots per 60 | 6.94 | 1 |
In overall production, Subban once again outclasses Karlsson, although this time by a much thinner margin. At even strength however, it’s Karlsson who is superior. Many analysts I know differ in their thinking about what’s more important for production from a defenseman, producing at even strength or on the powerplay. Most defensemen do their damage on the powerplay, but Karlsson’s offense first mentality allows him to act as a 4th forward often, and his even strength production seems sustainable. Whether you value overall or even strength production more in this case seems entirely subjective to me, but once again the two players are comparable.
| League Rank | Subban | Statistic | Karlsson | League Rank |
| 43 | 9.26% | Even strength on-ice shooting percentage | 9.58% | 14 |
| 7 | 18.18% | 5vs4 on-ice shooting percentage | 13.44% | 19 |
| 10 | 10.3% | Personal shooting percentage | 7.3% | 14 |
| 3 | 2.36 | 5vs4 goals per 60 | 0.61 | 30 |
| 1 | 8.25 | 5vs4 points per 60 | 4.44 | 10 |
For the powerplay, only players with over a minute of powerplay time per game and over 30 games played in 2013 are counted, while in 2011-12, 70 games is the cutoff point.
Subban has enjoyed better luck personally in his shooting, as well as on the powerplay, while getting the essentially the same luck as Karlsson at even strength. His increased PP luck though, does not cover the gap in powerplay production between the two players. Subban has been an absolute stud, whereas Karlsson is merely fantastic.
Part 2 – Role
The biggest complaint about Subban’s season has been that he isn’t playing tough minutes. But is that really true? A glance at quality of competition ratings tell that story, but I think it’s necessary to go deeper. I don’t think Corsi relative quality of competition tells an entirely truthful story about competition faced, especially in this short season. So instead I’m going to look at a variety of statistics to gauge each player’s competition level at even strength, attempting to separate out offensive talent from defensive talent in opponents.
| League Rank | Subban | Statistic | Karlsson | League Rank |
| Tied 18 | 50.8% | Opposition goals for percentage | 50.4% | Tied 37 |
| 20 | 2.319 | Opposition goals for per 60 | 2.352 | Tied 22 |
| Tied 60 | 50.0% | Opposition shots for percentage | 50.0% | Tied 63 |
| 17 | 29.259 | Opposition shots for per 60 | 28.875 | Tied 91 |
| Tied 84 | 49.8% | Opposition Fenwick for percentage | 49.9% | Tied 81 |
| 25 | 40.743 | Opposition Fenwick for per 60 | 39.984 | Tied 93 |
| Tied 85 | 49.8% | Opposition Corsi for percentage | 49.9% | Tied 83 |
| 22 | 55.257 | Opposition Corsi for per 60 | 54.048 | 76 |
What you can see in this table, is that in spite of all the people saying that Subban does not play tough minutes, in all 4 categories of possession he’s facing extremely potent offensive players. Ranking in the top 30 in every category means he’s clearly used as a number one defenseman, shutting down the opponent’s top producers. The players he plays against though, aren’t so sound defensively, and he’s been able to exploit that. The result is that in every measure, he faced nearly identical competition to Karlsson in possession, while ranking higher on the league scale in six of eight measures, most dramatically from a possession standpoint in goals for percentage.
| League Rank | Subban | Statistic | Karlsson | League Rank |
| 37 | 23.2 | Time on ice per game | 25.32 | 9 |
| 91 | 16.25 | Even strength time on ice per game | 19.86 | 5 |
| Tied 37 | 1.6 | Shorthanded time on ice per game | 0.55 | 210 |
| 38 | 52.8 | Offensive zone start percentage | 57.1% | Tied 8 |
In the time on ice battle, Karlsson clearly comes out way ahead. This is due almost entirely to Michel Therrien easing Subban in after his extended contract negotiations caused him to miss the first 6 games of the season, but even last year when Subban played the whole year, he played 2 minutes fewer at even strength per game than Karlsson did, and a minute less overall.
What we do know though, is that over Subban’s first 18 games, he averaged 14.9 minutes of even strength ice time per game, meaning the following 18 he averaged 17.6 minutes, which is closer to what’s expected. Also notable is that when Ellliott Friedman brought up Subban as a Norris candidate last Thursday on Hockey Night in Canada, Subban ranked 89th in ice time per game among defensemen, he’s climbed to 37th in just a couple of games.
The scales tip back in Subban’s favour when you look at the difficulty of minutes played, with Subban playing over a minute more per game in shorthanded situations and rising, as Subban has seen 2.14 minutes per game on the penalty kill in his last 18 games.
Subban also plays much tougher zone starts, albiet not as tough as he did last season. 52.8% in the offensive zone is by no means the toughest situation for a player of Subban’s talents, but contrast that with Karlsson, who was the 8th most zonally sheltered regularly used defenseman in the NHL last season, and it’s no contest. Subban may play fewer minutes, but they’re significantly tougher ones.
Part 3 – Defensive Partner
A lot was made last season of how Karlsson dragged a useless possession player around all year in Filip Kuba. But Subban’s regular partner is by no means a driver of possession either. We can use each defensive partner’s Corsi percentage with their star defenseman and without their star defenseman to see if Kuba is that much more of a drag on Karlsson than Josh Gorges is on Subban.
| Subban | Statistic | Karlsson |
| 56.1% | Gorges/Kuba’s Corsi percentage with Subban/Karlsson | 53.2% |
| 47.4% | Gorges/Kuba’s Corsi percentage without Subban/Karlsson | 44.4% |
| 58.6% | Subban/Karlsson’s Corsi percentage without Gorges/Kuba | 58.8% |
| +8.7% | Positive Corsi differential for Gorges/Kuba | +8.8% |
| -2.5% | Corsi drag on Subban/Karlsson | -5.6% |
Oddly enough, both star players prop up their main defensive partner by a nearly identical amount. Karlsson though, was dragged down by Kuba by over double the amount.
How to explain this? Well when Karlsson wasn’t paired with Kuba, he was playing next most often with Jared Cowen. Cowen is a positive possession player in his own right, though only marginally. He and Karlsson had significant chemistry, as both players played worse away from each other. The two of them together had a 56.3% Corsi percentage. He had a similar situation with his 3rd most frequent defense partner, Chris Phillips.
Subban on the other hand, nearly exclusively plays with Francis Bouillon when not with Gorges. Bouillon hasn’t been a possession disaster at 49.5% away from Subban, but together with Subban they have a 53% Corsi percentage. Subban away from Bouillon? 59.6%, a -6.6% drag on P.K..
I’m not saying that Subban is a better player than Karlsson, although it is my opinion that he’s a more complete player right now. Karlsson is a year younger, and a heck of a player. But Subban’s dominance can no longer be denied.
A lot of folks don’t want to admit it, be it for preconceived notions about P.K., not seeing him play night and night out, or other reasons, but this season Subban is putting up should make him an automatic for the Norris. It seems like the tide of public opinion is finally turning in favour of Subban, and unless something crazy happens in the last 6 regular season games, expect P.K. Subban to be the first Montreal Canadiens‘ name to be engraved on the James Norris Memorial Trophy since Chris Chelios in 1989.
