Let's get one thing out of the way. P.K. Subban is an extremely talented and a very dangerous player and the Ottawa Senators will have to figure out a way to contain him. Now is he as good as Erik Karlsson?
Erik Karlsson is very good. So much as that the good people at Habs Eyes on the Prize are using Karlsson's 2011-2012 season as the benchmark for whether Subban should win the 2013 Norris, and not his competition this year. I can see why Canadiens' fans would obsess over wishing Subban was at a Karlssonian level, but is simply not factually true.
Let's look at the basic statistics first. Subban was tied for first among NHL defencemen with Kris Letang at 38 points, although Subban's 42 games was seven more than Letang played. It should be noted that Letang played with far superior talent than Subban. The 0.90 points/game by Subban were second among defencemen, while Karlsson's 0.82 was third. So Subban was a little bit better than Karlsson offensively over half a season at first glance. Or was he?
Subban had 12 points on even strength all season while Karlsson had nine in just 17 games. Chris Phillips had 13 even strength points this season for Ottawa. But Subban thrived on the power play and has to be given credit for that. But are those kinds of numbers sustainable? The Canadiens had 329 minutes and 15 seconds of power play time this season, which is almost 30 minutes more than second-place Detroit. Subban himself had 197:50 of power play time this season despite playing only 42 games, which was fourth behind only teammate Andrei Markov, Alexander Ovechkin and Ilya Kovalchuk. The Canadiens were among the league leaders in power play time last season, but were 16th in the year before that.
Is Subban even a better power play player than Karlsson? He had a power play point for every 7.61 minute of power play time while Karlsson had one power play point for every 19.175 minutes. Even last year, Karlsson had a power play point for every 11.27 minutes of power play time. On the other hand, Subban mans the point with Andrei Markov who had 23 points on the power play himself. Last year when Karlsson had 28 points on the power play, the next closest defenceman on Ottawa had 14. The Canadiens are shooting 16% with Subban on the ice in 5-on-4 situations, while the Senators are shooting 7.69% with Karlsson. By comparison, Montreal shot 9.09% on the power play with Subban last year while Ottawa shot 13.44% with Karlsson. For reference, the league average on the power play is 13.5% this season. Montreal gets about 40 shots on net per 60 minutes of Subban's power play time and that number rises to 61 if missed shots are also counted. Ottawa gets 49 shots on net per 60 minutes of Karlsson power play time and that rises to 70 after adding missed shots.
In terms of 5-on-5, it is the same situation. The Senators shot 6.53% with Karlsson on the ice (9.58% last season) and Montreal shot 9.70% with Subban (9.83% last season) on the ice. Despite Karlsson's unfortunate luck with shooting % and/or talent of teammates, and Subban's good fortune, Karlsson was only slightly behind Subban in terms of points/game this season. Karlsson had 4.65 shots on goal per game this season which led all players in the NHL, Alexander Ovechkin was second at 4.58. No one else was even close with Rick Nash third at 4.00 and Dustin Byfuglien was second among defenceman at 3.3. Does Karlsson take long shots from the outside? Sure he does. His average shooting distance is 47.7 feet, but Subban's is 55.6 feet. Same thing on the power play, with Karlsson shooting from an average of 47.9 feet and Subban from 55.2 feet.
If Karlsson and Subban swapped on-ice 5-on-5 and 5-on-4 shooting percentages, Subban would have 21 points in 38 games while Karlsson would have 22 points in 17 games. This highlights how much the bounces just haven't gone for Karlsson offensively this season and for Ottawa in general.
There is no doubt, Karlsson is far superior to Subban in terms of offensive talent so let's look at defensive play. Karlsson is on the ice for 49 opposition shot attempts per 60 even-strength minutes while Subban is on the ice for 47. Counting only shots that hit the net, Karlsson is on for 30 shots against while Subban is on for 22.5. What this says is that both players spend an equal amount of time in the defensive zone. The difference in shots hitting the target is that Ottawa blocks 9.6 shots per 60 minutes with Karlsson on the ice, while Montreal blocks 15.6. Subban himself has blocked 49 shots (in all situations) in 42 games while Karlsson has blocked 14 shots in 17 games. This doesn't explain the large difference in blocked shots, so crediting Subban for his teammates blocking shots is likely not a valid representation of reality.
So is Erik Karlsson really not as good as P.K. Subban at preventing goals? We know from shot attempts that both players spend about an equal amount of time in the defensive zone. Karlsson has been a little luckier with his goaltender stopping 93.5% of shots with him on the ice while Subban's goaltenders have stopped 92.6% of shots. Despite this, Subban has been on for slightly fewer goals (1.66 versus 1.96). This difference may be due to the higher number of shots that Montreal blocks as a team, rather than any superior defensive skills by Subban.
Overall, per 60 minutes of Subban's 5-on-5 ice time, the Canadiens get 15.16 more shot attempts than the opposition does. With Subban off the ice, the Habs have an advantage of 1.46 shots, giving Subban a relative impact of 13.7 shots. With Karlsson on the ice, Ottawa had 22.28 more shot attempts than the opposition. With him off the ice, the Sens had an advantage of 2.23 giving Karlsson a relative impact of 20. This relative impact is what is called Corsi Relative. Again blocked shots are helping Subban in the offensive zone as well. The Senators have 21.6 shots blocked per 60 minutes of Karlsson while Subban has 17.7 of Montreal's shots blocked with him on the ice. Karlsson spends a lot more time in the offensive zone with 72 shots attempts compared with 62 for Subban.
Both are very good, and both teams have the puck a lot more than the opposition with the respective players on the ice. What kind of minutes do they play? Karlsson's opposition has an average Corsi Relative of 0.702, which is the third highest on the team among defencemen. Subban's opposition has an average Corsi Relative of 0.026, which is the third easiest among the seven defenceman to play 20 games. Karlsson starts in the offensive zone 58.3% of the time, while Subban starts in the offensive 53.6% of the time. Karlsson plays 21 minutes a night on even strength, while Subban plays 17 minutes a night on even strength. Karlsson is the most important player for the Senators and plays two minutes more than any other player club 5-on-5. Subban is fourth out of six defenceman on Montreal in even strength ice time per game, although he is only 37 seconds behind first.
To summarise, Karlsson leads all NHL players in shots per game, had to deal with injured teammates, bad luck with absurdly low team shooting percentage and still managed to put up similar numbers to Subban who has had more positive luck on his side. Karlsson also plays more and tougher minutes 5-on-5 although he does start in the offensive zone a little more than Subban.
P.K. Subban is a wonderful player who has dominated on the power play, with a little bit of luck. As good as Erik Karlsson though? I don't think so.