News Article: Jets have second most "Star Power" in NHL

[Yukon Joe]

http://www.thehockeynews.com/news/a...e-it-and-how-much-does-it-take-to-win-the-cup

It's The Hockey News, so nothing great, but it's the dog days of summer so I thought I'd post it. The writers thesis is that you need top-end "star players" in order to win championships. So he attempts to rank players by a measurement called Game Score, and see who had the most high-end talent.

The article itself never really talks about the Jets, but the Top teams were ranked as follows:

Nashville (6 stars, 22 points)
Winnipeg (6 stars, 21 points)
Pittsburgh (7 stars, 20 points)
Washington (7 starts, 19 points)
Calgary (6 stars, 19 points)
Boston (4 stars, 19 points)
Chicago (5 stars, 18 points)
Tampa (4 stars, 15 points)
Ottawa (4 stars, 14 points)
St Louis (4 stars, 14 points)
Columbus (4 stars, 14 points)

And according to this guy's analysis, teams with 20+ points have gone on to the final 4 just under 50% of the time, and have wont the Cup ~15% of the time.

I'm not completely sold, but I was just struck at how the list of the top teams made sense and included all the suspected playoff teams, but Winnipeg was a big fat outlier having not made the playoffs but sitting in #2 spot.

Oh, for the curious our "star players" were Wheeler (1st Tier), Schiefele, Byfuglien (2nd Tier), Laine, Trouba (3rd Tier) and Ehlers (4th Tier)

Click to read more...

 

[Gm0ney]

There's some discussion about this over in the Around the League stickied thread.

Also, .906 goaltending has been the main problem with this team for 5 out of the last 6 years. A team's not going anywhere with that.

 

[Rheged]

There's a bit of a running gag around how highly his game score model has ranked the Jets in the past so this isn't really that surprising haha.

It kinda just re-enforces a lot of the sentiment around here that goaltending and to some extent coaching have held back what has the makings of a pretty good team.

 

[Snot Rocket]

[mod]

...and back to said article, I would like to see where Laine will fit in after this year.

 

[Whileee]

4 of the Jets "stars" have just reached that level in the past season or two. That combined with sub-par goaltending explains why they haven't made a run yet, but are poised to do so in the next year or two.

 

[StatisticsAddict99]

Not fond of the rankings... allot of Mis-placed players...

 

[Eyeseeing]

http://www.thehockeynews.com/news/a...e-it-and-how-much-does-it-take-to-win-the-cup

It's The Hockey News, so nothing great, but it's the dog days of summer so I thought I'd post it. The writers thesis is that you need top-end "star players" in order to win championships. So he attempts to rank players by a measurement called Game Score, and see who had the most high-end talent.

The article itself never really talks about the Jets, but the Top teams were ranked as follows:

Nashville (6 stars, 22 points)
Winnipeg (6 stars, 21 points)
Pittsburgh (7 stars, 20 points)
Washington (7 starts, 19 points)
Calgary (6 stars, 19 points)
Boston (4 stars, 19 points)
Chicago (5 stars, 18 points)
Tampa (4 stars, 15 points)
Ottawa (4 stars, 14 points)
St Louis (4 stars, 14 points)
Columbus (4 stars, 14 points)

And according to this guy's analysis, teams with 20+ points have gone on to the final 4 just under 50% of the time, and have wont the Cup ~15% of the time.

I'm not completely sold, but I was just struck at how the list of the top teams made sense and included all the suspected playoff teams, but Winnipeg was a big fat outlier having not made the playoffs but sitting in #2 spot.

Oh, for the curious our "star players" were Wheeler (1st Tier), Schiefele, Byfuglien (2nd Tier), Laine, Trouba (3rd Tier) and Ehlers (4th Tier)

We have a few very good underrated players.
The one guy that gets little love and lots of skepticism is Wheels on the main boards.
He's a top 5 RW and almost nobody outside Winnipeg even knows that.

 

[SoCalJetsFan]

We have a few very good underrated players.
The one guy that gets little love and lots of skepticism is Wheels on the main boards.
He's a top 5 RW and almost nobody outside Winnipeg even knows that.

That is to be expected since he has played all of his best hockey in a small market team that has never made any noise in the playoffs. Now with Laine on the team he could continue to fly under the radar with fans from other team no matter how successful the Jets are. However, you can bet that Wheels doesn't fly under anybody's radar that has to play against him.

As to the article, I think we should expect league average goal tending this season and if the Jets can't make any noise this year well then there is only one other area to look at. I won't mention any names but their first initials start with Paul, Charlie, Wade.

 

[Eyeseeing]

That is to be expected since he has played all of his best hockey in a small market team that has never made any noise in the playoffs. Now with Laine on the team he could continue to fly under the radar with fans from other team no matter how successful the Jets are. However, you can bet that Wheels doesn't fly under anybody's radar that has to play against him.

As to the article, I think we should expect league average goal tending this season and if the Jets can't make any noise this year well then there is only one other area to look at. I won't mention any names but their first initials start with Paul, Charlie, Wade.

I'd love nothing more to silence some [...] on the main boards.
We have some very good players and with the 2 good additions will finish well above 6th in the Central where the average TML fans are placing us.
I'm also a TML fan but a small but vocal minority are pretty disrespectful towards us

 

[Babooch]

He's a top 5 RW and almost nobody outside Winnipeg even knows that.

Agree, and I think it's because he produces like a star but often doesn't look like one. Not uncommon for him to stumble and bumble his way through his first few shifts yet end up with a point or two.

I look back to 2.0 year one when he had a measly 2 assists after 17 games how bad his first impression was with many of us. It took awhile for me to like him.

 

[Jet]

Not found of the rankings... allot of Mis-placed players...

They lost me at Byfuglien - 2nd tier

PS I used my 22,000th post to bash Buff! Staying on message haha

 

[raideralex99]

Weird a team with star power can't make the playoffs. What does that tell you players are not the problem?

 

[Whileee]

Weird a team with star power can't make the playoffs. What does that tell you players are not the problem?

As noted above, 4 of the 6 just moved into "star" territory within the past year or two. The window for contention really starts now.

That, and goaltending.

 

[heretik27]

So according to this guys chart, Nylander is tier 3, but Marner is tier 5? Bryan Little doesn't register in those top 5 tiers? I just quickly skimmed it, but I'm already feeling it's pretty much bologna.

 

[truck]

So according to this guys chart, Nylander is tier 3, but Marner is tier 5? Bryan Little doesn't register in those top 5 tiers? I just quickly skimmed it, but I'm already feeling it's pretty much bologna.

I really like it. Not a perfect measure - as he states in the article, but a decent snap shot. His metrics said the Jets had 6 first line forwards last year.

 

[Zhamnov5GoalGame]

So according to this guys chart, Nylander is tier 3, but Marner is tier 5? Bryan Little doesn't register in those top 5 tiers? I just quickly skimmed it, but I'm already feeling it's pretty much bologna.

Someone on the main boards mentioned that is was a good concept but that the rankings were way off.
Malkin in Tier 2????
I couldn't be bothered to click the link to read it after that.

 

[garret9]

To quote myself on Twitter:

Every model is imperfect and therefore wrong sometimes. Every human being is imperfect, and therefore wrong sometimes.
When critiquing a model vs another, it is essential to test them. One of the most important factors is: which one is wrong less often.
Now, looking at the exceptions is important when looking on how to improve the previous methods. This is why/how we move forward.
How do you test how often something is wrong? The sniff test can be useful, but keep in mind your sniff test will also be wrong too.
If your sniff test was perfect, than stats wouldn't be necessary. Part of the value of models is finding things you'd get wrong.
So, when you wish to critique models, you need to critique it on the methodology, the background theory, and the results from testing.

This is a mistake many have made when criticizing this article.
Yes, the order is not perfect, and no order ever will be. That's not how you can critique a model though.

Now, an example of how you can critique well...

Game score is basically a combination of a player's scoring, individual shots, EV Corsi +/-, and EV Goal +/-, predominately in that order. This means the statistic is prone to skew from usage effects, especially teammates and team quality, but also somewhat from ice time as well.

To compare against WAR, another holistic statistic, Game Score has one large disadvantage: measuring defensive impact.

When we moved from Corsi, to xGoals, to XPM (a major component of WAR), the largest area in improvement we saw was gaining a better grasp of a player's defensive impact. The improvement in defensive impact was huge.

This makes sense when you re-look at the list as a trend instead of individual situations. There definitely is a trend towards highly valuing offensive contributors.

Game Score is still very much useful.

It does a very good job in measuring offensive impact, and Corsi & Goal differentials makes there still be some defensive/two-way impact measure there (although GS doesn't weight Corsi or GD very highly). I would actually suggest Game Score as being an extremely useful tool for fantasy hockey leagues. The scoring component is obviously useful for fantasy leagues, but the Corsi and GD components help with +/- in fantasy leagues.

I've also used Game Score to supplement BPM, the other important input to WAR, when BPM is not available. BPM measures impact a player has on the game through boxscore stats, which is predominately scoring and taking shots. BPM is pretty much Game Score adjusted for usage with Corsi and Goal Differentials removed.

 

[Guardian17]

I'll believe we have a good team when we make the playoffs consistently.

 

[psycho_dad*]

They lost me at Byfuglien - 2nd tier

PS I used my 22,000th post to bash Buff! Staying on message haha

Yeah I actually laughed out loud. Buff. What exact criteria are they using?

 

[garret9]

Yeah I actually laughed out loud. Buff. What exact criteria are they using?

Maybe read their article to find out.

It's mostly points, with a bit of shot differentials, goal differentials, and penalty differentials, looking at the past three seasons, weighted with last season being the highest.