Thursday, March 29, 2007

The Rod Langway Award

I've always liked the idea of an annual trophy for the NHL's top defensive defenceman.

After all, the one end-of-season award available for blueliners, the Norris Trophy, long ago became dedicated to offence-first types, leaving the league's underrated defensive stalwarts going unrecognized. (The last player to win the award with fewer than 50 points was Rod Langway in 1984.)

But being a top defensive player is a tough thing to qualify, whether it's a forward or a defenceman. As someone who followed the Vancouver Canucks closely in the past, for instance, I've always thought Mattias Ohlund was, if not the team's best player, certainly one of them.

And that's not really something that shows up on his statline as a player that has never scored more than 36 points and has been a minus nearly as many years as a plus in the plus/minus column.

The past few days, I've been looking into coming up with a simple way to find who the league's top defensive blueliners have been this season, the combination of a few key metrics that will give us an idea of who has stood out in the category.

In order to come up with a manageable number of blueliners to look at, I came up with a few criterion:
  1. They'd have to play a lot, let's say be among the top 90 defencemen in even-strength ice time
  2. Playing shorthanded would also be a big part of being a shutdown blueliner, so we'll take the top 60 from that category
Those two stipulations leave us with 52 defencemen, which is a workable number to use. With that group, I've chosen three metrics to look at (with another hat tip to Behind The Net):

(a) rate of goals against at even strength,
(b) rate of goals against shorthanded and
(c) strength of opposition

Here's a look at the leaders when all three categories are weighted equally:

RK Player
Team GP +/- EVGA
/60
rank SHGA
/60
rank EV Opp rank
1 JOHNSSON
MIN 73 -4 2.23 9 3.10 2 0.0685 16
2 PHILLIPS
OTT 77 28 1.76 1 5.16 15 0.0842 12
3SCHULTZ
MIN77-22.31134.3150.076315
4 HANNAN
SJS 74 -1 2.28 11 4.11 4 0.0572 18
5
LIDSTROM
DET 76 36 1.82 2 5.85 24 0.0897 9
6
MARKOV
MTL 72 3 2.51 22 4.53 8 0.0950 7
7
VOLCHNKV
OTT 74 32 1.90 4 6.46 32 0.0957 6
8
O'DONNELL
ANA 74 10 2.29 12 4.36 7 0.0218 35
9
HAMHUIS
NSH 76 7 2.72 31 5.56 19 0.1068 4
10
ROZSIVAL
NYR 75 5 1.89 3 3.48 3 -0.0203 49
11 BIEKSA
VAN 76 -1 2.43 15 5.25 16 0.0456 25
12 OHLUND
VAN 74 -5 2.43 15 2.83 1 0.0119 41
13 VLASIC
SJS 76 14 2.43 15 4.99 13 0.0325 31
14 FERENCE
BOS 76 7 2.09 7 4.32 6 -0.0147 47
15 ALLEN
FLA 77 6 1.93 6 5.42 17 0.0172 38
16 WITT
NYI 75 15 2.53 24 6.25 29 0.0892 10
17 REGEHR
CGY 74 23 1.90 4 7.08 42 0.0542 19
18 SARICH
TBL 77 -10 2.43 15 6.46 33 0.0643 17
19 SEABROOK
CHI 75 -6 2.46 20 4.94 12 0.0284 34
20
KLESLA
CBJ 69 -11 2.62 28 6.06 27 0.0868 11
21 SMITH
EDM 77 -10 2.92 39 5.87 25 0.1386 2


For one thing, the list shows how inaccurate a measure plus minus can be — at least if you put any stock in these rankings.

Kim Johnsson's certainly a surprise at the top of the list, but given he plays the most minutes per game on the top defensive team in the league, perhaps he shouldn't be. Johnsson's not getting easy minutes either, and plays a ton on one of the best penalty kills in the league, so perhaps it's time we gave him his due as a solid defensive player.

There are some other surprises there that highlight players who have had terrific seasons in the role: Dan Hamhuis, Bryan Allen, Marc-Eduoard Vlasic, Brendan Witt and Brent Seabrook.

As for the bottom end of the 52-player list I've put together, it offers a telling story the other way:

22 BEAUCHEMIN
ANA 66 7 2.37 14 5.56 20 0.0300 33
23 KOMISAREK
MTL 77 8 2.54 25 6.33 30 0.0837 13
24 NIEDERMAYER
ANA 74 8 2.57 26 5.67 22 0.0538 21
25 BREWER
STL 76 -9 2.94 41 6.55 34 0.1378 3
26 MORRISONN
WSH 73 3 2.76 34 6.91 40 0.0976 5
27 SUTER
NSH 77 13 2.46 20 4.68 10 -0.0216 51
28 ZHITNIK
ATL 74 1 2.98 42 4.58 9 0.0338 30
29 SCUDERI
PIT 74 6 2.45 19 7.04 41 0.0484 24
30 BOUWMEESTER
FLA 77 23 3.03 45 4.80 11 0.0346 29
31 POTI
NYI 72 1 2.17 8 6.40 31 -0.0151 48
32 WARD
BOS 74 -4 2.61 27 5.83 23 0.0213 37
33 HILL
NYI 75 7 2.72 31 6.85 38 0.0542 19
34 KEITH
CHI 76 -2 2.93 40 5.65 21 0.0348 27
35 CHARA
BOS 74 -16 2.98 42 6.87 39 0.0933 8
36 VAN RYN
FLA 73 -6 2.65 30 5.48 18 0.0067 43
37 HAINSEY
CBJ 74 -14 2.51 22 6.24 28 -0.0131 46
38 COMMODORE
CAR 77 6 2.99 44 5.91 26 0.0347 28
39 NUMMINEN
BUF 74 13 2.62 28 6.71 37 0.0215 36
40 KALININ
BUF 76 11 2.73 33 6.69 36 0.0302 32
41 HAVELID
ATL 72 -2 3.17 50 8.53 50 0.1895 1
42 HATCHER
PHI 76 -24 3.68 52 5.12 14 0.0143 40
43 NORSTROM
DAL 70 -16 3.03 45 8.52 49 0.0789 14
44 BLAKE
LAK 72 -26 3.15 49 8.26 46 0.0702 15
45 STUART
CGY 69 -15 2.89 38 8.41 48 0.0456 25
46 CLARK
COL 76 4 3.05 47 7.57 43 0.0510 22
47 MCCABE
TOR 76 9 2.26 10 8.57 51 -0.0272 52
48 GILL
TOR 76 9 2.83 36 6.65 35 -0.0097 44
49 MILLER
LAK 77 -13 2.81 35 8.15 45 0.0103 42
50 MORRIS
PHX 76 -18 3.42 51 8.73 52 0.0487 23
51 KABERLE
TOR 68 3 2.83 36 8.37 47 -0.0104 45
52 MICHALEK
PHX 76 -15 3.14 48 7.62 44 0.0169 39


Keep in mind that offensive statistics count for absolutely nothing in these standings — only defensive performance is rated. Atlanta's Niclas Havelid takes a pounding in the goals against categories, but benefits from a big bump up due to the fact he faces the toughest opposition in the league, and Derian Hatcher is saved by the fact he's actually been very strong on the penalty kill this season in huge minutes there.

I'll be the first to admit this is a pretty rudimentary look at defensive performance among blueliners, but I'm hoping this can be a jumping off point for other bloggers/stats junkies to come up with their own winners for the Rod Langway Award.

UPDATE Just an FYI: Chris Pronger would have been in a top five position had he not missed 16 games due to injury. His goals against rates are way up there.

24 Comments:

At 1:37 AM, March 30, 2007, Anonymous David Johnson said...

You will notice that most of the defensemen on the first list play on teams with pretty good goalies (Minnesota, Ottawa, San Jose, Detroit, Vancouver, Nashville, etc.) and most of the defensemen on the second list play on teams with weaker goalies Phoenix, LA, Toronto etc. You also aren't accounting for time on the ice or who they are playing with. There are just too many factors that go into giving up goals that no simple statistical analysis will work.

I have attempted to create a rating system which I am somewhat happy with though I am still not correctly adjusting for goaltender influence. I plan on working on that this summer.

 
At 1:39 AM, March 30, 2007, Anonymous David Johnson said...

That link should be this.

 
At 1:48 AM, March 30, 2007, Blogger James Mirtle said...

Name a netminder who has won the Vezina while on a non-playoff team?

The fact is, players who have a solid supporting cast are often those who shine, and always those who win the end-of-season awards. I think this list offers a good look at which defencemen would be pretty good candidates for the nonexistent Langway Award.

This does account for time on the ice as the goals against ratios are based on ice time.

Although I realize my mini-analysis reflects poorly on your beloved Leafs (funny, that).

 
At 2:09 AM, March 30, 2007, Blogger El Brucio said...

There does need to be better defensive-minded stats. But I do object to the perception that the Norris is for "offensive minded" d-men. Nik Lidstrom, as you can see, is 5th on your list. Combine that with his PP presence and passing ability, and he's the best OVERALL d-man in the league. If it went based on offensive stats, you'd think Sergei Gonchar would've gotten some serious consideration and maybe finish 2nd or 3rd a couple of times.

But a quick look at your list isn't that surprising. Marc-Edouard Vlasic has been San Jose's most reliable blueliner at 19, and it's really unfortunate that he won't be a Calder finalist.

Andrei Markov and Mike Komisarek are a severely underrated defensive tandem, it's the rest of the D that allowed all those even strength goals. Kim Johnsson was really used well by Hitchcock pre-lockout, it's not surprising to see him at the top of this list in Minny, even if earlier in the year he didn't look that great on a new team.

 
At 9:09 AM, March 30, 2007, Blogger James Mirtle said...

There does need to be better defensive-minded stats. But I do object to the perception that the Norris is for "offensive minded" d-men. Nik Lidstrom, as you can see, is 5th on your list.

The only thing I wonder with regards to that is, why, other than Lidstrom, will none of the top 20 listed here get any Norris consideration at all?

Niedermayer's middle of the pack on a terrific team, and he'll almost assuredly finish as the runner-up to Lidstrom.

You're right — defence is taking into account when voting for the Norris; I just don't think it's done enough.

 
At 9:10 AM, March 30, 2007, Anonymous David Johnson said...

Although I realize my mini-analysis reflects poorly on your beloved Leafs (funny, that).

It is funny because that porous Leaf defense that everyone talks about has actually given up fewer shots on goal than Minnesota, and Ottawa, and New Jersey and Nashville and Calgary and 22 teams in total.

The main problem with your metric I think is the short handed goals. I believe, and I hope to try to prove it at some point, that short handed success has significantly more to do with goaltending and opposition than with the players who play the PK. Your first 12 players play on teams in the top 12 on the PK and with good goalies. 14 of the bottom 15 players play on teams with weak penalty kills and more often than not weak goaltending.

I guess the bottom line is that any metric which has an offensive minded defenseman with one of the worst +/-'s on his team as the top defensive defenseman in the NHL is highly suspect. Maybe a simple fix would be to weight even strength goals against double or triple the weight of short handed goals against. That would probably give you a more realistic result and would make sense considering the defensemen will play far more time even strength as short handed.

 
At 10:12 AM, March 30, 2007, Blogger James Mirtle said...

Good points.

My thinking with the shorthanded situations was that, if you're going to measure defensive performance, how better than to look at a situation which is entirely defensive in nature?

There's certainly room for improvement there, though.

 
At 10:17 AM, March 30, 2007, Blogger James Mirtle said...

And Kim Johnsson hasn't been much of an offensive defenceman this season with the Wild.

 
At 10:27 AM, March 30, 2007, Anonymous Anonymous said...

Kim Johnsson's certainly a surprise at the top of the list, but given he plays the most minutes per game on the top defensive team in the league, perhaps he shouldn't be. - Then again it proves how good he must be playing. Defensive teams, by their nature, play more in their own end and require the players to be better defensively since there is no margin for error.

 
At 10:28 AM, March 30, 2007, Blogger teebs said...

One thing that sticks out to me is the fact that if you combined this with the overall best offensive defensemen Lidstrom would be at the top of the pile on both lists.

So to me that says that the Norris has been going to the right player in the last few years.

As for others on this list, I do think that defensive ability does factor in to the choices fairly heavily. Niedermayer, Pronger and Lidstrom all just happen to be such special players who can play almost 30 minutes a game in any situation excel defensively and put up nearly a point a game. In any sport what so ever you choose those types of players to single out and recognize as the best players in the game.

And as a Canucks fan I would obviously love to see Mattias get some recognition for his fantastic shut down game against some of the top players in the league. This has been somewhat of an off year for him. Alot of his shut down time has gone to Mitchell and Bieksa, and his offensive numbers seem to be matched by Salo and Bieksa.

Actually I would think that Mitchell would be a lot higher on the list if he was hurt less this season as well.

 
At 10:32 AM, March 30, 2007, Anonymous Anonymous said...

I agree with your premise sine there is already an award dedicated to defensive forward. However, since Langway did win the Norris doesn't that disprove your theory?

The voting goes in waves, sometimes the offensive guy gets it (Coffey), sometimes the defensive guy (Langway), but mostly the right guy (Lidstrom, Bourque, Pronger).

 
At 10:34 AM, March 30, 2007, Blogger James Mirtle said...

So... because one defensive defenceman won it 23 years ago that means all blueliners of that ilk are getting recognition?

It's not even so much winning the Norris — look at the voting. There won't be one defensive player (like Phillips) in the top 10, which isn't a true representation of the 'top defenceman' designation.

 
At 10:42 AM, March 30, 2007, Blogger Jes Gőlbez said...

*sigh*

The myth that the Norris has become about offensive defenseman is bogus. You never see guys like Sandis Ozolinsh, Bryan Berard, or Sergei Gonchar get any real votes, do ya?

The big difference between a guy like Lidstrom and Mitchell is that Lidstrom can pile up the points ALONG with great defense. There are quite a few good defensive d-men in the game, but it takes a special talent to combine a high level of offense as well.

Guys like Pronger, Niedermayer, Lidstrom, et al are not poor defensively and are certainly near the top of their craft.

 
At 11:27 AM, March 30, 2007, Blogger James Mirtle said...

Is anyone in this conversation saying Lidstrom's poor defensively?

 
At 11:46 AM, March 30, 2007, Blogger Jeff J said...

I think a separate award for defensve defensemen is a great idea. Only it should be the Eddie Shore award, not the Langway.

 
At 11:48 AM, March 30, 2007, Blogger JavaGeek said...

Scoring rates against can be very deceiving. Just look at Ryan Smyth.

I would be willing to argue Smyth was Edmonton's best penalty killer this season, but he had a 5.4GA/Hr rating, which looks bad compared to Pisani's 3.5GA/Hr. However since Smyth spent a lot of time (71%) on the PK with Horcoff (7.2GA/Hr) his GA average is inflated. Smyth is currently sitting at 2GA/Hr in New York.

You still haven't considered how the evil 5 on 3s will effect the best players PK score.

 
At 11:49 AM, March 30, 2007, Blogger Jeff J said...

A. Markov is #6. Hmmm... didn't I see a reference on this blog not too long ago about Montreal's top two defensemen being "offensive-minded?" ;)

 
At 11:53 AM, March 30, 2007, Blogger James Mirtle said...

Yeah, you're right Jeff — and that's why I'm starting to do some of this stuff. There are so many stereotypes about certain players that aren't backed up by anything statistical, and these lists point out guys you'd never think of in this context...

Just look at Rusty Klesla. Or Bryan Allen, who really deserves recognition for how well he's played.

 
At 1:29 PM, March 30, 2007, Anonymous twain said...

Why not give the Norris to the best defenceman period, as it currently attempts to do, and make a new award for offensive defenceman--The Bobby Orr award (who, of course, played great defense too, but that is not the point).

It would eliminate the offensive contributions bias, and retain an award for the best D-man full stop.

 
At 1:51 PM, March 30, 2007, Anonymous Anonymous said...

"So... because one defensive defenceman won it 23 years ago that means all blueliners of that ilk are getting recognition?"

No. But the reverse also holds true as well. Only one "Offensive" defencemen has won it in the past twenty years as well (Coffey as sited). Your opening premise is that the Norris goes to the best offensive defensmen this just isn't true. This is a different argument than saying you want a trophy for best defensive defensmen just because they deserve recognition. That is why I agree on your point for a "Langway" award, but not on your characterization of the Norris.

The Norris is actually pretty good at deciding the best "overall" defencemen. THat means good at both ends of the ice. If a player is so dominant in only one aspect they occasionally win the Norris.

Chara is by no means known for his offense, but did get a few votes last year and conversely McCAbe didn't.

 
At 4:56 PM, March 30, 2007, Anonymous ken said...

I think the "Langway award" should go to the best defenceman on the worst team.

Think how that guy feels, when he works his butt off to knock the puck off Crosby's stick, only to see his buffoon teammates give it right back. He needs cheering up.

 
At 3:10 AM, March 31, 2007, Anonymous vadim sharifijanov said...

"You never see guys like Sandis Ozolinsh, Bryan Berard, or Sergei Gonchar get any real votes, do ya?"

ozolinsh finished second one year. gonchar has finished in the top four too, as have other one-dimensional offensive guys like mccabe, iafrate, and housley. chris chelios, a stellar defensive defenceman and a defence-first guy for much of his career, finished with over 70 points three times in his career. those were the only seasons he won the norris. the closest scott stevens came to winning was the year he almost scored 20 goals and 80 points, by far his career highs in both. and derian hatcher was a finalist during his career offensive year. like james said, it's not that nobody considers defensive prowess when voting for the norris, only that too much weight is given to offence.

 
At 6:52 PM, March 31, 2007, Blogger Hawerchuk said...

I've been off email for two days (how is that even possible?) so I'm a little late to this discussion.

Anyways, I think GA/60 is too strongly-affected by the quality of your goalie. Johnsson matches up against the other team's first line, but the Wild do pretty poorly when he's on the ice. I need to spend a while running regressions on the strength of opposition to see how much that really affects performance.

David - looking forward to your new and improved stats...

 
At 3:17 PM, April 01, 2007, Blogger PDO said...

"I would be willing to argue Smyth was Edmonton's best penalty killer this season, but he had a 5.4GA/Hr rating, which looks bad compared to Pisani's 3.5GA/Hr. However since Smyth spent a lot of time (71%) on the PK with Horcoff (7.2GA/Hr) his GA average is inflated. Smyth is currently sitting at 2GA/Hr in New York.

You still haven't considered how the evil 5 on 3s will effect the best players PK score."

Kinda funny that you make that second point, being Horcoff has likely played more 5 on 3 time than any other Oiler forward this season ;)

 

Post a Comment

<< Home

Links to this post:

Create a Link


.

Free Page Rank Checker
eXTReMe Tracker