NHL.com Trophy Tracker - Hart Trophy

[Duffalufagus]

Lol people seriously think Matthews is going to win the Hart? Okay.

Vegas does. But you probably know more than Vegas.

 

[hockeyisgud]

I think voters are starting to get a little wise, McDavid and Dri are minute hogs, usually they play all 2 minutes of the PP and have gotten burnt many times by being gassed, If Gaudreau or anyone else is 10 points out I think they win.

Imagine Gaudreau's point totals if he played 4 more minutes a game mostly on the PP, his +/- would most likely suffer but his point totals wouldn't.

It doesn't increase in a linear fashion for obvious reasons but that has been beaten to death so I won't explain it again

 

[Sidney the Kidney]

Dom isn’t wrong. Gaudreau easily has a better Hart case than Huberdeau.

Wasn't the issue that Dom's model had all three of Gaudreau, Lindholm and Tkachuk as more Hart-worthy because it (the model) couldn't differentiate between the three players that play on one line?

The other issue was that he didn't even consider Huberdeau's season worthy of a top 10 finish. At the time, Huberdeau was either 1st or 2nd in league scoring. IMO, that's just silly.

 

[Duffalufagus]

Right, but you said that voters no longer just sort by points because they've "gotten smarter". There's actually no evidence to suggest that's the case, even if you only want to use the past 3 seasons. I mean, Draisaitl won the Hart only 2 seasons ago despite having relatively poor underlying numbers (compared to the other finalists), largely because he ran away with the scoring race.

Just because the odd voter like that Dom guy will base his entire voting card on who has the best GSVA and nothing else, doesn't mean that's widespread among the majority of voters. If it was, you'd have seen many more "analytics darlings" winning the Hart versus "raw point total darlings".

Agree. However, one can only hope the writers will eventually look at (and try to understand) the numbers in more depth. They really speak to who plays the most complete 200 ft game.

 

[tarheelhockey]

Wasn't the issue that Dom's model had all three of Gaudreau, Lindholm and Tkachuk as more Hart-worthy because it (the model) couldn't differentiate between the three players that play on one line?

Yep, and the irony is rather thick when you consider that the whole point of the model is to make that type of distinction. Part of the reason these analytical approaches gained traction in the first place was that people got fed up with +/- when it… failed to differentiate between linemates.

The other issue was that he didn't even consider Huberdeau's season worthy of a top 10 finish. At the time, Huberdeau was either 1st or 2nd in league scoring. IMO, that's just silly.

That’s a more controversial take, but I don’t think it’s a silly one. There are 2-3 goalies and 2-3 defensemen who have been worthy of consideration. Has Huberdeau been a top 4-6 most valuable forward this season? It’s fair to say yes, but I would say no and a lot of other people would too.

Even taking it for granted that Huberdeau belongs in that top 4-6, in no way shape or form has he been driving the success of the Panthers the way Gaudreau has driven the Flames. I mean there’s just no way. And IMO Gaudreau has been the context-neutral better player regardless of the point totals.

 

[Erik Alfredsson]

Vegas does. But you probably know more than Vegas.

Yeah I do know more than Vegas actually.

 

[Sidney the Kidney]

Yep, and the irony is rather thick when you consider that the whole point of the model is to make that type of distinction. Part of the reason these analytical approaches gained traction in the first place was that people got fed up with +/- when it… failed to differentiate between linemates.

These models seem to have just re-branded plus/minus in a new package, using things like shot attempts and scoring chances instead of goals for and against, and are trying to paint them as being more accurate at determining value.

This is why I'm still skeptical of them. They've still got a tendency to be too much a measure of what all 5 guys on the ice at one time do and not nearly enough a measure of what a particular player on the ice at that time does.

That’s a more controversial take, but I don’t think it’s a silly one. There are 2-3 goalies and 2-3 defensemen who have been worthy of consideration. Has Huberdeau been a top 4-6 most valuable forward this season? It’s fair to say yes, but I would say no and a lot of other people would too.

Even taking it for granted that Huberdeau belongs in that top 4-6, in no way shape or form has he been driving the success of the Panthers the way Gaudreau has driven the Flames. I mean there’s just no way. And IMO Gaudreau has been the context-neutral better player regardless of the point totals.

I'm not arguing Huberdeau over Gaudreau, specifically. But at the time of that article, I don't see 10 players more deserving of a nomination than Huberdeau.

 

[Hockey Outsider]

Wasn't the issue that Dom's model had all three of Gaudreau, Lindholm and Tkachuk as more Hart-worthy because it (the model) couldn't differentiate between the three players that play on one line?

That's a common issue with analytics models.

The most comprehensive model that I've seen so far is Alan Ryder's Point Allocation - a very technical read, but interesting if you're into this kind of thing.

To list a few examples:

• In 2003, the system ranked Milan Hejduk over his Hart-winning linemate Peter Forsberg. The system read his goals as being more valuable than Forsberg's assists (even though it was obvious who was driving the play). The system also ranked them as virtually even defensively, although it was again Forsberg driving that.
• In 2006, the system ranked Joe Thornton only about 6% ahead of Jonathan Cheechoo (who was tied for 4th among forwards, ahead of Jagr, Alfredsson, Staal, Crosby, and others). The system couldn't "tell" that it was Thornton driving the play. Thornton lost out on being the top forward in the model as a result.
• In 2009, the system ranked Alexander Semin tied for 5th in the league among forwards (ahead of Art Ross winner Malkin, plus Iginla among others). Some of Ovechkin's success was likely misattributed to Semin (and this allowed Datsyuk to rise to the top of the best forwards list instead of Ovechkin).

My point isn't to criticize Ryder's model (like I said, it's the gold standard for hockey analytics, as far as I'm concerned). But any model will have trouble differentiating between linemates, even if it's obvious to fans who's driving the play. People who construct models should acknowledge this, but often don't (because why detract from your model's credibility if you're trying to get more subscribers?)

 

[tarheelhockey]

These models seem to have just re-branded plus/minus in a new package, using things like shot attempts and scoring chances instead of goals for and against, and are trying to paint them as being more accurate at determining value.

This is why I'm still skeptical of them. They've still got a tendency to be too much a measure of what all 5 guys on the ice at one time do and not nearly enough a measure of what a particular player on the ice at that time does.

That’s exactly what they are. On a very fundamental level, nearly the only “events” which can be objectively measured in hockey are shots, the locations of shots, the outcome of shots, etc. Almost anything else requires some level of personal judgment/bias.

Upside: shot attempts and scoring chances are much more common than goals (by orders of magnitude) so even if that is the only improvement, it is a very real improvement on +/-in terms of sample size.

Downside: algorithms that derive solely from shot-oriented data have all the same issues as +/- in other areas, especially failing to determine the causes of their outcomes.

This is why NHL teams ignore publicly available data, and instead hire trustworthy professionals to custom-design data based on video review and according to internal standards. One guy (Dom) simply cannot do this for every single NHL game while also writing articles for a national audience every other day.

 

[thaman8765678]

It has to be Shesterkin or Gaudreau.

 

[BonMorrison]

Josi.

 

[george14]

Dom isn’t wrong. Gaudreau easily has a better Hart case than Huberdeau.

Why do people insist on making these statements? Huberdeau is a finalist right now, so is Gaudreau most likely. They are both in the same ballpark along with like 5 other dudes.

 

[authentic]

I think voters are starting to get a little wise, McDavid and Dri are minute hogs, usually they play all 2 minutes of the PP and have gotten burnt many times by being gassed, If Gaudreau or anyone else is 10 points out I think they win.

Imagine Gaudreau's point totals if he played 4 more minutes a game mostly on the PP, his +/- would most likely suffer but his point totals wouldn't.

If Gaudreau got the same powerplay time he would easily be the point leader right now.

 

[authentic]

Why do people insist on making these statements? Huberdeau is a finalist right now, so is Gaudreau most likely. They are both in the same ballpark along with like 5 other dudes.

Huberdeau has no case for the Hart. He's not even the most valuable player on his team, and Barkov won't get a nomination either.

 

[tarheelhockey]

Why do people insist on making these statements? Huberdeaubis a finalist right now, so is Gaudreau most likely. They are both in the same ballpark along with like 5 other dudes.

Andersen is probably a Vezina finalist, in the same ballpark with like 5 other dudes.

I insist on making the statement that Shesterkin easily has a better Vezina argument than Andersen. Being a finalist in the ballpark with some other dudes is not an argument for #1.

And I don’t think Huberdeau’s a finalist right now unless we’re just translating Art Ross rankings over to Hart votes. There’s no good argument that he’s been more valuable to his team than Shesterkin, Matthews, Josi, Gaudreau, Makar, even Ovechkin or McDavid. Not too many teams have shaken their heads and said “wow, we would have had the Panthers if only we could’ve figured out a solution for Huberdeau”.

 

[Mez]

Josi.

81 points now.. leading his team. Also great defensively...he has to get top consideration.

 

[Roof Daddy]

If Josi keeps up his torrid pace, how can it be anyone but him? Especially when most had the Preds picked as a fringe playoff team, likely on the outside looking in. He’s put the team on his back.

I think it’s kinda odd we haven’t seen a D win the Hart since Pronger.

 

[Soundgarden]

If Josi keeps up his torrid pace, how can it be anyone but him? Especially when most had the Preds picked as a fringe playoff team, likely on the outside looking in. He’s put the team on his back.

I think it’s kinda odd we haven’t seen a D win the Hart since Pronger.

Yeah, most everybody and their mother's picked Nashville to be only above Arizona and maybe Chicago in the west back in the off season.

 

[duul]

I think an obvious metric for 'most valuable' is whoever is at the top of the goal share for his team. Like highest point contribution % to overall team offence.

Every year it's McDavid followed by Draisaitl is it not? How are they not the most valuable forwards for sure, when removing them hinders your group offensively more than any other players?

Defence and goalie are their own thing, but in terms of forwards it should be clearly obvious. Here are some numbers for player share of team offence:

Aho - 11.8%
Gaudreau - 11.5%
Mackinnon - 11.8%
Marchard - 15.5%
Huberdeau - 12.3%
Barkov - 11.7%
Josi - 8.5%
Crosby - 12%
Matthews - 12.7%

McDavid - 21.3%

One of these is not like the other.

 

[EXTRAS]

Josi needs some votes

 

[MessierII]

I would have shesterkin Josi and McDavid for the finalists.

 

[george14]

Andersen is probably a Vezina finalist, in the same ballpark with like 5 other dudes.

I insist on making the statement that Shesterkin easily has a better Vezina argument than Andersen. Being a finalist in the ballpark with some other dudes is not an argument for #1.

And I don’t think Huberdeau’s a finalist right now unless we’re just translating Art Ross rankings over to Hart votes. There’s no good argument that he’s been more valuable to his team than Shesterkin, Matthews, Josi, Gaudreau, Makar, even Ovechkin or McDavid. Not too many teams have shaken their heads and said “wow, we would have had the Panthers if only we could’ve figured out a solution for Huberdeau”.

Before we start endorsing TarHeelHockey as the official NHL award expert, maybe let’s wait until the season ends? Enough with the hyperbole and anecdotes; are you speaking to NHL players?

Shesterkin had a couple of bad starts and the gap closed mightily. Notice how he was the MVP front-runner? That changed within a week.

Huburdeau has a 25 point lead on Barkov, leads the league in assists, and is 3rd in points. Personally I think he is a finalist along with Matthews and Gaudreau. However, that will change I’m sure because the voters will have a hell of a time choosing. Josi, Shesterkin, McDavid, Kaprizov, etc. Hell, Crosby would be a finalist on my ballot had he not missed time.

 

[WillTheThrill]

I wanna see Matthews win it so bad just to see this board implode.

 

[Frosty415]

I would have shesterkin Josi and McDavid for the finalists.

My finalists too

 

[tarheelhockey]

Before we start endorsing TarHeelHockey as the official NHL award expert, maybe let’s wait until the season ends? Enough with the hyperbole and anecdotes; are you speaking to NHL players?

Shesterkin had a couple of bad starts and the gap closed mightily. Notice how he was the MVP front-runner? That changed within a week.

Huburdeau has a 25 point lead on Barkov, leads the league in assists, and is 3rd in points. Personally I think he is a finalist along with Matthews and Gaudreau. However, that will change I’m sure because the voters will have a hell of a time choosing. Josi, Shesterkin, McDavid, Kaprizov, etc. Hell, Crosby would be a finalist on my ballot had he not missed time.

What part of sentences starting with “I think…” implies I did anything other than express an opinion? No need to re-frame it as anything other than that, just to undermine it without addressing the points made.

Shesterkin had a couple of bad starts, sure. In fact he had three starts of .879 or less…

… in his past 18 games. He actually had five starts of .970 or more during the same timeframe. If three off-games is enough to tank an otherwise Hart-worthy season, what do we do with the 11 games where Huberdeau didn’t register a point? Common sense says we look at those games in context and realize they don’t mean much on an 82-game schedule.

If Shesterkin tanks the final month of the season, sure, write him off. But that doesn’t mean that Huberdeau had the better argument when the argument was made.

As of now, Huberdeau:

- doesn’t have an argument based on points — he’s 4th, behind McDavid/Draisaitl/Gaudreau

- doesn’t have an argument based on “carries his team” — he’s plainly behind guys like Shesterkin/Gaudreau/Josi and arguably not really distinguishable from Kyle Connor for that matter.

- doesn’t have an argument based on “straight-up best player” — not even in that conversation

Yes, he’s the top scorer on a good team. That’s the argument. Good for him, he’s having a fantastic season, a career season, possibly setting up for a Cup run. That’s not a Hart argument of any substance, not even compared to a weak winners like Taylor Hall (who was at least demonstrably the best player on his team by a large margin).