View Single Post
05-18-2008, 03:31 AM
I voted for Kodos
Sturminator's Avatar
Join Date: Feb 2002
Location: West Egg, New York
Country: Ukraine
Posts: 8,612
vCash: 500
Originally Posted by Hockey Outsider View Post
Great work, Sturm. I was planning on doing the same type of analysis with the Norris votes, you beat me to it.

I think you did the right thing by limiting this to years where players received at least one first place vote or two top-3 votes. It's obvious that in many years, all of the players outside of the top 5-10 simply received one or two votes (out of 50-100 potential voters). There's no point in pretending that a player with a single 4th place vote is definitely superior to a player that received a single 5th place vote (or didn't receive any votes at all).

A few quick observations. Randy Carlyle's poor voting record supports my long-held opinion that he's the worst Norris winner ever. Zubov's voting record is pretty interesting- it supports the general consensus that he's been a very good defenseman for a long time, but he's only really had a few elite seasons. Finally, I think that Phaneuf has sky-high potential but I'm very surprised by his three straight top-eight finishes. Maybe I should watch more Flames games.

There's one more Norris metric I'd like to work on: Norris shares. It's based on the Vezina shares Dr. No has on his website. Basically, you compare voting points a player has, to the total available votes. This way, Bobby Orr's unanimous 1970 Norris is worth more than, say, Doug Wilson getting ~58% of available points.
Thanks, HO. The shares issue is something that I thought of while I was working on the project, but didn't have the energy for at the time. I think it would be rather illuminating, but what share shall we compare: overall votes, 1st place votes or possible points for a single player?

I think possible points is the best metric available, though it will not yield results that add up to a clean 100% across all players - only a comparison of 1st place votes yields that kind of clean data. Given that a shares metric will mostly only be relevant for analyzing the top performers in any given season, perhaps 1st place votes is the way to go. I dunno.

I'm a little surprised by Phaneuf's Norris voting record, as well, though maybe I shouldn't be. I must admit to a some surprise over the consistency shown in Rob Blake's Norris voting record, as well. I'm a guy who has criticized Blake in the past for his brain farts and inconsistency, and yet, viewed from this angle, Blake actually appears to have been quite a consistent performer over his peak years, and perhaps unfairly maligned. Ditto Pronger, whose ups and downs do play out in the voting, but who we're reminded has had a lot more up than down, especially when injuries are taken into account.

Sturminator is offline   Reply With Quote