Who Deserves the Norris Trophy?

The NHL award that seems to be generating the most interest and discussion this season is the Norris Trophy given to the best defenseman. In some circles Drew Doughty seems to be the favourite because he is having a great season in a great career on a team having a great year. Others are picking Erik Karlsson because of his dominant offensive season for a defenseman, one that few defensemen can match in the history of the game. Brent Burns is having almost as good of an offensive season but for a playoff team so his name gets thrown into the mix too. Kris Letang is having an outstanding second half of the season putting him into the mix as well.

Along with Doughty, Karlsson, Burns and Letang I am going to throw Victor Hedman, Ryan Suter, Shea Weber, Duncan Keith and John Klingberg into the discussion because they are who I have seen mentioned in Norris discussion or who I feel deserves some consideration. I am sure I am leaving a few notables out but I am also certain the Norris Trophy will and should come from this group of players.

All statistics in this article, unless stated otherwise, are 5v5 even strength statistics.

Offensive Performance

Norris_PointsPer60

The easiest measure of offensive production is straight up point scoring rates. Here Karlsson is clearly king but Burns is not too far behind with a higher goal rate but significantly lower assist rate. Hedman, Klingberg and Letang all look pretty strong here as well.

On the flip side, this is an area where Doughty’s Norris trophy credentials are definitely lacking.

Another measure of offensive production is GF60 RelTM which attempts to measure a players impact on the offensive production of the players he plays with.

Norris_GF60RelTM

Karlsson still leads here but Hedman is almost his equal. Drew Doughty looks a little better than his raw point rate which is a result of the fact that he plays on a team that struggles to score at 5v5. You can’t rack up points if your team doesn’t score many goals.

Although point totals generate the wow factor GF60 RelTM is probably a better measure of offensive contribution.

Defensive Performance

Norris_GA60RelTM

GA60 RelTM attempts to measure the players contribution to goal prevention. Here lower numbers are better as it indicates the players has an ability to reduce goals against rates. This was Drew Doughty’s strength this season while the offensive defensemen Karlsson, Burns and Letang performed quite poorly.

Now I know some people are going to scream “But players can’t influence save percentage, you should look at CA60 RelTM” but I completely disagree. I do not believe it is coincidence that Karlsson, Burns and (to a lesser extent) Letang – the three highest scoring defensemen in the league this year – are among the worst defenders in GA60 RelTM over the last 3 seasons. Weber is having an atypical season for him but  his role seems to have changed to a less defensive oriented role. The last two seasons he has been on the ice for 51.8% and 50.9% of the Predators defensive zone faceoffs. This season that has dropped to 43.9%. I have written enough on this topic so I won’t go into it further here. Besides, the Norris Trophy goes to the best *performance* by a defenseman, not most talented (that is a somewhat different question to answer).

Zone Starts and Roles

The following chart shows how coaches have used each of these Norris trophy candidates.

Norris_OZFO_vs_DZFO_Roles

The majority of defensmen have been assigned a higher percentage of his teams offensive zone starts than defensive zone starts with Karlsson being given the most significant offensive role. The exceptions are Shea Weber who had a slightly higher percentage of defensive zone faceoffs and Duncan Keith who was given a significant defensive role.

The zone starts don’t directly impact a players statistics all that much but zone starts can indicate player role which is an indication of what the coach expects of the player. Here we can probably conclude that Duncan Keith is expected to prevent goals as much or more than he is expected to produce offensively. As a result his offensive production may be inhibited by the role he is given. Karlsson conversely is expected to produce more offense (which he does well) and asked to play defense less. This is important because if we evaluate defensemen based primarily on their offensive statistics without taking roles into account we unfairly penalize defensemen who are asked to play a more defensive shut down role.

Overall Performance

The following chart summarizes the above by plotting GF60 RelTM vs GA60 RelTM.

Norris_GA60RelTM_vs_GF60RelTM

For me this chart is what has me leaning towards Hedman as my Norris trophy pick. Klingberg is interesting as well but he probably should be eliminated from discussion due to his lower quality of competition (not discussed in this article) and lack of PK usage (also not discussed in this article). His low face off deployment likely also indicates the coach doesn’t trust him for a specific role or line matching duties which is probably typical for young defensemen.

Beyond Hedman you have Karlsson for pure offensive ability and Doughty for stellar defensive performance and Duncan Keith deserves consideration for putting up solid numbers despite by far the toughest defensive role. I know people are wowed by Karlsson’s numbers but I personally have a problem awarding the Norris Trophy to a defenseman who plays like a forward while his team suffers defensively while he is on the ice.

In conclusion, my pick for Norris Trophy is Victor Hedman. Runner Ups are Doughty, Keith and Karlsson in that order.

 

This article has 12 Comments

  1. It seems you have two numbers for Shea Weber’s defensive zone face offs this year. Or did I misunderstand that?

  2. So basically you pick Hedman because of his position on your little chart. Great hockey analysis *eyeroll*

    1. Thanks for the feedback. Next season I’ll create charts of the statistics that I think are the most important in evaluating defenseman but instead of choosing the player who I evaluate as the best I’ll pick a name out of a hat.

  3. Good analysis, David. As a Tampa fan I was pretty surprised to see Hedman get your pick. He’s been good this year, but not nearly as good as last year. However, wouldn’t his lack of powerplay output count against him? He certainly is a terrific 5 v 5 player, but can’t gain steady PP minutes and in the time he does play his production is not great.

    Shouldn’t a Norris winner contribute more in high leverage situations like PP?

    1. IPP of at 5v4 play:

      Doughty: 88.5
      Karlsson: 82.8
      Hedman: 75.0
      Letang: 69.7
      Weber: 68.8
      Burns: 65.0
      Suter 56.7
      Keith: 56.5
      Klingberg: 55.9

      So as a percentage of the points available to him on the PP Hedman trails on Doughty and Karlsson. How much do we penalize Hedman for playing on a team with one of the worst performing power plays in the league? How much is he to blame for that? Some maybe but for me not enough to change my conclusion.

  4. While I agree that defenseman can impact save percentage I thinking looking at once ice goals for/against is going to great overstate this effect.

    Karlsson was 3rd in Corsi this season and combined with his fantastic offensive performance that more than makes him a solid candidate, the one who I think is most deserving of the title.

    If you were making this case last season when his CA60 was significantly worse than average I would have agreed with you (though interestingly enough his GA60 was better last year that this year).

    While I agree the selectors greatly overrate point production I just disagree that that alone us enough to exclude Karlsson.

    1. The problem with Corsi is that if you start in the OZ more often, you’ll have a higher Corsi #. It doesn’t take into effect that if your team wins the OZ faceoff, you’ll get a shot for more than likely or if your team loses a DZ faceoff, you’ll give up a shot. Karlsson gets a very high OZ start, so he gets more shots for than against which skews his Corsi to positive. What is really bad, is how high his corsi is, yet his goal against is not better. That to me means he usually gives up very high scoring chances (as a result of him trying so hard to score).

  5. What about players like OEL, TJ Brodie, Stralman, Muzzin, Giordano or Subban? Are they worth considering?

    1. I am a big OEL fan and he probably deserves some consideration. Brodie and Giordano are both good and you could inlude one of them but I don’t think this year was one of their better years. Stralman is good but second best on his team so shouldn’t be included. I am a huge fan of Subban as well but he is in the Karlsson/Burns category – good offense, less impressive defensive numbers. Because Karlsson and Burns had such great offensive seasons they deserve consideration over Subban. Of the three I would probably choose Subban to build a team around though. I think Muzzin is overrated by the hockey analytics community and does not deserve to be in Norris consideration.

      Another guy I like is Pietrangelo and probably deserved some consideration.

  6. Hi David,

    I think your use of Rel TM stats in this case is somewhat problematic, as a player on a worse team will be able to have a greater impact on GF Rel TM as well as GA Rel TM. For example, Doughty is playing on LA, which has been a solid defensive team for years now (and was again this season in terms of GA), and his GA Rel TM numbers are still extremely impressive. I believe that this undervalues him, even though he came out on top in terms of defensive value. Additionally, although you didn’t pick him as your candidate for the Norris, the fact that Ottawa is a poor team makes it easier for Karlsson to have a higher GF Rel TM.

    Maybe you took this into consideration when making your picks, and to me it seems that way based on your runner up selections, but the graph at the end does not seem to me to be a good enough indicator of a defenceman’s performance over the entire season.

Comments are closed.