Top Centers by WAR Heading Into the 2019-20 Season

[NHL WAR]

Round 2. I'll start off this time by clarifying that the specific rankings aren't the be all end all. They are made just using WAR. It is a 1.0 template list that could be further refined with the eye test and additional data.

That being said, here is the top 10:
McDavid
Crosby
Stamkos
Matthews
Tavares
Barkov
MacKinnon
Bergeron
Point
Malkin

Order may a bit off, but the players are pretty spot on. Point is the only one that didn't make the top 10 in the poll a couple weeks back.

Here is the full list:
The Top 31 Centers by WAR Entering the 2019-20 NHL Season

Click to read more...

 

[Portable Mink]

Now do top centres by beard length

 

[Khelandros]

WAR doesn't work in hockey. No one player can shape a game like they can in baseball, maybe a goalie. If one player could change a game, the Oilers would have won the cup 5 years running.

 

[Chileiceman]

WAR doesn't work in hockey. No one player can shape a game like they can in baseball, maybe a goalie. If one player could change a game, the Oilers would have won the cup 5 years running.

Look how well the Amgels are doing with Mike Trout...

 

[izzy]

Now do top centres by beard length

top 10 most muscular legs

 

[Dr Jan Itor]

Seems like the right players are there.

 

[canuckking1]

Look how well the Amgels are doing with Mike Trout...

Maybe he meant basketball

 

[Krewe]

Point is the only one that didn't make the top 10 in the poll a couple weeks back.

Still don't understand how Stamkos was voted top 10 but not Point. There is a reason Stamkos got demoted to 2C. Stamkos puts up big numbers but Point is better now after last season (I am aware your model uses a larger sample size)

 

[Breakers]

WAR doesn't work in hockey. No one player can shape a game like they can in baseball, maybe a goalie. If one player could change a game, the Oilers would have won the cup 5 years running.

Do they have WAR in basketball?
I feel it would be really relevant there because one guy in basketball can really really shape the game.
I've seen Lebron drag rotting corpes into the playoffs.

 

[Yackiberg8]

Look how well the Amgels are doing with Mike Trout...

That theory was shut down pretty quick.

 

[IPS]

I got Brian Boyle at the top.

He'd have been a beast in Vietnam.

 

[TomasHertlsRooster]

WAR doesn't work in hockey. No one player can shape a game like they can in baseball, maybe a goalie. If one player could change a game, the Oilers would have won the cup 5 years running.

I don’t agree with this. I think that the WAR models that exist today in hockey aren’t nearly as comprehensive or accurate as the WAR models that exist in baseball, but they’re still a useful tool that provides value and they’re worth looking at.

 

[Leonardo87]

WAR! What is it good for?

 

[Khelandros]

WAR! What is it good for?

Absolutely nothing!

 

[JoVel]

WAR! What is it good for?

Increasing domestic manufacturing.

 

[Burke the Legend]

Kind of a fail article, the methodology section doesn't even discuss how the WAR score is calculated. We may be able to get some kind of reasonable WAR score eventually in hockey (although it still wouldn't be as close to precise as baseball) but the stats guys don't have all the right tracking tools yet as the tech is still not there yet, plus there's a number of competitive models with regards to calculations.

BTW Big fan of Brayden Point but I'm not sure if his 21.5 shooting % from last year is sustainable.

 

[NotProkofievian]

How do you have passing data?

 

[Rebels57]

Nice to see Sean "not a top 50 player according to TSN" Couturier ranked as the 14th best Center in WAR. Then again, he's top 10-20 in pretty much every metric you can reasonably use to rate a forward.

 

[Trap Jesus]

Did you take into account the missed games for Bergeron? I think you did that for Hall in the last one you did.

 

[ToDavid]

WAR doesn't work in hockey. No one player can shape a game like they can in baseball, maybe a goalie. If one player could change a game, the Oilers would have won the cup 5 years running.

Saying it "doesn't work" is a bit of an oversimplification. It may never be the be-all-end-all stat like it can be in other sports but it can be a useful metric as long as you understand its limitations.

 

[Filthy Dangles]

WAR doesn't work in hockey. No one player can shape a game like they can in baseball, maybe a goalie. If one player could change a game, the Oilers would have won the cup 5 years running.

And one player can make more of a difference in baseball?

Value over Replacement Player stats work more in baseball because it's a more individual sport. You can't be carried or elevated much offensively or defensively like you can be by a linemate or teammate in hockey. For offensive players, it's just you vs a major league pitcher repeated a bunch of times and those results can be compared to other hitter vs pitcher matchups fared.

 

[Nineteen67]

WAR doesn't work in hockey. No one player can shape a game like they can in baseball, maybe a goalie. If one player could change a game, the Oilers would have won the cup 5 years running.

Maybe you need you more than just one guy, I don’t know.

 

[Gary Nylund]

I don’t agree with this. I think that the WAR models that exist today in hockey aren’t nearly as comprehensive or accurate as the WAR models that exist in baseball, but they’re still a useful tool that provides value and they’re worth looking at.

Agreed and that list looks pretty good too. I mean people will never agree 100% on anything but if 9 out of the 10 on this list match the poll recently taken here, it's hard to really complain about this. Is there any other stat that would come this close to matching the players picked in the poll? I doubt it.

 

[lomiller1]

What WAR model are you using, because there is a lot of variation between them. Something to keep in mind is that (AFAIK) none of the commonly available WAR models do particularly well as predictive metrics. IOW they work better for describing what a player did than predicting what they will do. The @DTMAH model showed good results in out of sample tests, but it hasn’t been publically available for a few years now.

 

[NHL WAR]

How do you have passing data?

If you mean pass tracking data, I don't. The passing component in my formula is assists and the impact a player has on the quality and quantity of their teammate's shots.