Tuesday, April 08, 2008

Picking the Selke winner

The Frank J. Selke Trophy has to be the most miss-voted upon trophy in all of professional sports. There's absolutely no rhyme or reason to some of the ballots entered, year after year, and the winners are almost never the top shutdown players in any given season.

That's not to say everyone gets it wrong.

Here's a sampling of some of the candidates that are out there this season from various sources:
Danny Heatley and Daniel Alfredsson both are great on both ends of the ice, with good +/- rankings, and would be good choices. But they also will split votes, and the guy to me that stands out tied Alfredsson in short-handed goals and assists this year on a non-play-off team: Patrick Sharp. The other finalists will and should probably include Alfredsson, and I am going to go with Brendan Morrow, a physical forward with a strong +/-.
Pavel Datsyuk, Detroit. A plus-41 rating should start the conversation, but he wins 54.4 percent of his face-offs and led the NHL with 142 takeaways. Runner-up: Ryan Kesler, Vancouver.
Talk to anybody in the Red Wings locker room, you'll hear there's no better two-way forward than Henrik Zetterberg. Everybody sees the obvious offense — 43 goals and 92 points — but he's a plus-30, brilliant in the neutral zone and on the penalty kill. Runners-up: Mikko Koivu of the Wild; Saku Koivu of Montreal, John Madden of New Jersey, Chris Drury of the N.Y. Rangers.
There's some good sniffing going on there, with many of the right candidates coming out. It's just painful that plus-minus is the only measure available to weigh in on the discussion.

Except it's not.

According to statistics such as strength of competition, goals against per minute played and penalty killing metrics, Mikko Koivu has been the NHL's top defensive forward this season. And Anaheim's checking line continues to be deadly.

Here's my list of the top 30 shut down performers (normally I wouldn't even show plus-minus here, but I will to prove a point):

1 MIKKO KOIVU MIN 57 0.15 1.55 4.53 15
2 TRAVIS MOEN ANA 77 0.16 1.35 5.61 -9
3 ROB NIEDERMAYER ANA 78 0.17 1.42 6.24 -6
4 SAMUEL PAHLSSON ANA 56 0.18 1.71 6.19 -6
5 PAVEL DATSYUK DET 81 0.09 1.86 5.71 43
6 KRIS DRAPER DET 65 0.03 1.81 4.39 -4
7 JAY PANDOLFO N.J 54 0.16 1.58 6.50 9
8 DANIEL CLEARY DET 62 0.05 1.95 5.15 18
9 STEPHANE VEILLEUX MIN 77 0.10 2.37 5.10 -13
10 RICK NASH CBJ 80 0.08 2.40 4.90 -1
11 PATRICK SHARP CHI 78 0.08 2.79 3.28 10
12 PATRICK RISSMILLER S.J 79 0.03 2.20 4.17 -9
13 MICHAEL PECA CBJ 65 0.06 2.24 5.18 -3
14 HENRIK ZETTERBERG DET 74 0.13 1.91 6.60 34
15 CRAIG CONROY CGY 79 0.07 2.23 5.43 10
16 JEFF HALPERN T.B 83 0.06 2.31 5.18 0
17 MIKE MODANO DAL 82 0.08 2.87 3.36 -9
18 P.J. AXELSSON BOS 75 0.09 1.89 6.61 9
19 PAUL STASTNY COL 66 0.06 2.49 5.00 26
20 MARCO STURM BOS 80 0.06 1.86 6.54 8
21 MATHIAS TJARNQVIST PHX 78 0.03 1.85 5.74 -2
22 TORREY MITCHELL S.J 82 0.05 1.92 6.26 -4
23 ALEX BURROWS VAN 82 0.06 1.82 6.84 6
24 JOHAN FRANZEN DET 71 0.05 1.34 7.03 13
25 JOHN MADDEN N.J 80 0.13 2.25 6.57 5
26 MIKE GRIER S.J 78 0.05 2.49 5.15 -14
27 STU BARNES DAL 79 0.01 2.33 3.61 -9
28 JORDAN STAAL PIT 82 0.03 1.96 5.94 -1
29 CHRIS DRURY NYR 82 0.01 2.17 4.90 1
30 STEVE OTT DAL 73 0.09 2.02 7.48 0

If you could give the award to a whole line, that'd make things rather easy.

Now, Koivu has missed a ton of games, about one-third of the season, and so has Pahlsson, another good candidate. And the one intangible I haven't included here is offensive production on the penalty kill, which has to at least count for something.

And, in my mind anyway, that means Patrick Sharp deserves this one for 2007-08. (Although any of the top 15 are better choices than the majority of winners in previous years.)


Labels: ,


At 8:10 a.m., April 08, 2008, Anonymous Anonymous said...

Great remark there about Selke, I have wondered the same thing, why do they stare at that +/- department so tightly.

ps. First time on this site, good stuff, keep it up. I will be back

At 8:50 a.m., April 08, 2008, Blogger Jeff J said...

Other things to consider:

SHTOI - should a player be considered if he's not one of the coach's first choices for the PK?

Weighting SHGA and EVGA - It's tempting to be impressed by Sharp's low SHGA/60, but you have to remember that even for these guys PK time is usually less than 20% of their total TOI. For Sharp it's less than 10%. When counting goals against, you have to give their EV numbers 80%+ of the weight.

Goalies - Playing in front of good goaltending gives you lower numbers.

I think I'd pick PJ Axelsson.

At 8:53 a.m., April 08, 2008, Anonymous Anonymous said...

No, you can't give it to the entire line, but you can easily give it to their leader in the faceoff circle, Sammy Pahlsson.

At 11:14 a.m., April 08, 2008, Blogger Daniel said...

Pahlsson was shafted last year, without a doubt. But I am not sure he gets it this year. For all the Ducks games I have watched this, he didn't stand out like he did last year, he was incredible. He still is, I would bet that if you questioned the Western conference stars, they would all say they hate going against Samuel Pahlsson. So maybe he does get it and maybe that should be how they give the award away, ask the opposing players.

At 11:48 a.m., April 08, 2008, Blogger nebcanuck said...

The reason people look at +/- is clearly because of the trophy's description. It goes to the best two-way forward, not the best defensive forward. In most people's mind, that includes offensive output. A guy who can get 100 points in a season and still have a reasonably low goals against per minute is a shoe-in to win over a player who gets 10 points but has top standing in regards to goals against.

That's in theory, at least. The truth is, a lot of years it goes to the guy with the best defensive skills, regardless of offensive output. Thus, I think +/- becomes a less important indicator. But if you're looking at the offensive direction as well, you have to rule out guys who only get a handful of points in a season, really.

At 11:53 a.m., April 08, 2008, Blogger nebcanuck said...

Or sorry, I should say the colloquial description. I guess it really is top defensive forward, but you never hear it tossed around that way.

At 12:16 p.m., April 08, 2008, Anonymous Anonymous said...

Yeah, +/- is problematic I suppose... Bob Gainey used to win the Selke while having a terrible +/- (relative to the rest of the Habs).

If Lafleur was +70 and Gainey was +10, wouldn't that mean that Lafleur was the far better 2-way player? (e.g. he'd be more likely to come out ahead, GF-GA wise.) Of course, Guy probably played as much as 25% more even strength time, which would inflate his +/- relative to Gainey.

Or should the Selke be for whoever is the best defensive player should the puck be in your defensive zone? (Even then, Lafleur might have been more likely to take the puck out of the def zone & go up ice & create offense.)

The Selke is probably the hardest trophy to award properly.

At 12:27 p.m., April 08, 2008, Blogger Earl Sleek said...

The truth is, a lot of years it goes to the guy with the best defensive skills, regardless of offensive output.

I think this is the most untrue comment in this thread. I've actually found it to be the opposite in recent history.

At 12:30 p.m., April 08, 2008, Blogger James Mirtle said...

Offensive production has very little to do with this award, although the voters have often disagreed.

At 2:09 p.m., April 08, 2008, Anonymous Anonymous said...

The actual NHL criterion for the Selke: among forwards who kind of have a "defensive reputation," who had the most media exposure this season?

One thing to note: because a good offense is a good defense in life as well as hockey, technically guys that have a good puck possession game and are always on the attack are, in the most literal sense, the best "defensive forwards" to have. But that seems to kind of kill the spirit of the award, doesn't it?

At 2:26 p.m., April 08, 2008, Blogger Steve said...

But...but...Rod Brind'amour

At 4:55 p.m., April 08, 2008, Anonymous Peter Drake said...

GA60 is a very misleading stat for an individual. It really represents the defensive ability of a line, plus the goalie.

When considering Alfredsson, his linemates are bigtime gamblers but they score more goals than they allow. Yet somehow those goals against are supposed to indicate that Alfredsson isn't a great defensive forward? He's a dominant penalty killer, scores shorthanded goals galore, and is the primary forward on the 3-on-5 PK.

Here a suggestion for a stats guru. Calculate the effect a player has on his teammates' net goals for each situation (EV, PK, PP, PP+2, PK-2, EN etc.), and then weight it based on the amount of time they're on the ice with the various teammates, normalized for the team's overall GA60, and adjusted for the quality of competition. That would tell you how much a skater contributes to beating the opponent. Consider adjusting the PK metric to give an extra reward for shorthanded goals, they're big momentum changers. If you did that you'd have a good representation of how good a player is at defending, relative to the rest of their team.

At 5:43 p.m., April 08, 2008, Blogger James Mirtle said...

Alfredsson doesn't just have a mediocre GA/60; he's right at the bottom of the list among players who meet the ice time requirements.

Ottawa's been one of the worst defensive teams in the league this season.


Post a Comment

<< Home

Links to this post:

Create a Link


Free Page Rank Checker
eXTReMe Tracker