Hockey Outsider
Registered User
- Jan 16, 2005
- 9,056
- 13,987
I love this and how much your list has evolved into a flowchart. I still think that of the more recent finishes, the Sedin and Perry picks consecutively baffle me to this day. Perry in particular.
First, I want to say that there's some really good work in this thread, including the flow chart. That said, the whole flow chart thing bothers me, but not in any way that is the OP's fault. It just bothers me that the voters' thought process can be boiled down to something like that. I'm not saying that those factors can't be used to help make the decision, because they certainly can, but too much of a reliance on them, which it feels to me that many voters have causes some problems, imo.
For the record, I don't take the flowchart too seriously. But there are some pretty obvious patterns in how the voters pick the winners, and I wanted to see how well it would hold up. Obviously I had to manufacture some of rules (such as Hall's win), and there's no guarantee it will hold up in the future. But it's somewhat disheartened that nearly two decades of Hart trophy winners can be identified with such a simple chart. It raises the question - how much value are the award voters really adding if their choices are so predictable?
(I'm not sure if anyone is familiar with Darryl Shilling - he had a fantastic website 15+ years ago with all types of cutting-edge (for the time) analytics work. He said he had, but never published, a Hart trophy prediction formula. He spot-checked it on 10-15 years worth of data from roughly 1950 to 2004, and he found it had about a 90% accuracy rate).
I'm sure Pronger's 2000 win messes with predictions (and is why you started in 2001). A case of no particularly strong seasons by either forwards or goalies plus an above average quality Norris season plus a media narrative that developed about how they had been underrating defensemen forever - harder to fit that unique year into a flow chart.
For the record, I don't take the flowchart too seriously. But there are some pretty obvious patterns in how the voters pick the winners, and I wanted to see how well it would hold up. Obviously I had to manufacture some of rules (such as Hall's win), and there's no guarantee it will hold up in the future. But it's somewhat disheartened that nearly two decades of Hart trophy winners can be identified with such a simple chart. It raises the question - how much value are the award voters really adding if their choices are so predictable?
(I'm not sure if anyone is familiar with Darryl Shilling - he had a fantastic website 15+ years ago with all types of cutting-edge (for the time) analytics work. He said he had, but never published, a Hart trophy prediction formula. He spot-checked it on 10-15 years worth of data from roughly 1950 to 2004, and he found it had about a 90% accuracy rate).