By the numbers: How the Jets skaters performed

[garret9]

There was a short interview clip somewhere here on HF. I imagine it was on the Oilers' board.

I posted it on our adv stats thread too.

Also, I just checked myself since you inspired me (also known as giving me something to distract me with hahaha)

R2 for guys who take lots of shot attempts and Corsi% is .2
R2 for past Corsi and future Corsi is about .8

When you take in mind that top 6 (aka more talented) players tend to take more shots on average, then that .2 is pretty telling that there is no real relationship.

 

[pucka lucka]

I posted it on our adv stats thread too.

Also, I just checked myself since you inspired me (also known as giving me something to distract me with hahaha)

R2 for guys who take lots of shot attempts and Corsi% is .2
R2 for past Corsi and future Corsi is about .8

When you take in mind that top 6 (aka more talented) players tend to take more shots on average, then that .2 is pretty telling that there is no real relationship.

So if I am interpreting you correctly, we should have made some trades yesterday.

 

[garret9]

lol not exactly sure...
depends on the move... I would say numbers speak for a very slim chance of Jets in playoffs

 

[garret9]

I wanted to take a look and see how these numbers compared to the last 3 seasons combined, since I already have the data-mined with me starting to look at UFAs:

Let me know what you think about these ones. Looks like multi-years give a bit better handle of things, although you lose who is trending up or down.

 

[Gil Fisher]

I'd be interested in seeing how Kane/Ladd compare to the best left wingers in the league.

 

[Jimby]

So, if I am reading this right Buff is our best offensive defenseman and Ellerby is our best defensive and best overall?

I wanted to take a look and see how these numbers compared to the last 3 seasons combined, since I already have the data-mined with me starting to look at UFAs:

Let me know what you think about these ones. Looks like multi-years give a bit better handle of things, although you lose who is trending up or down.

 

[YWGinYYZ]

BZed - those numbers are affected / coloured by usage and QoC as well - Ellerby receives much more 'cushy' minutes than someone like Buff, Bogo or Enstrom, who regularly see more d-zone starts and are up against higher QoC opponents.

 

[Jimby]

isn't that supposed to be compensated for with the delta corsi? A positive delta corsi would be above the average expected results of all NHL defensemen?

BZed - those numbers are affected / coloured by usage and QoC as well - Ellerby receives much more 'cushy' minutes than someone like Buff, Bogo or Enstrom, who regularly see more d-zone starts and are up against higher QoC opponents.

 

[garret9]

BZed - those numbers are affected / coloured by usage and QoC as well - Ellerby receives much more 'cushy' minutes than someone like Buff, Bogo or Enstrom, who regularly see more d-zone starts and are up against higher QoC opponents.

It is a regression using a players TOI, the opponents avg corsi, your teammates average corsi, and zone starts.

So the effects of those factors should be extremely diminished.

BUT I've noticed from the regressions that team effects still seem to be slightly under regressed (only IMO not statistically founded), so guys on super elite possession teams like LAK all have good numbers. Guys on super terrible possession teams like EDM all have bad numbers.

 

[garret9]

I'd be interested in seeing how Kane/Ladd compare to the best left wingers in the league.

Kane and Ladd are really good possession players (Ladd more so than Kane)

 

[ps241]

Really cool read and I did comment

As I said on site I have issues with judging possession with my raw eye and I have viewer bias issues.

 

[garret9]

Really cool read and I did comment

As I said on site I have issues with judging possession with my raw eye and I have viewer bias issues.

I'll reply here because you're likely to see it here first haha.

This isn't as simple as judging posession but rather possession relative to their peers + usage. You'd have to be very good hockey scout to judge that. I know I wouldn't see all that hahaha.

Ladd has been pretty consistently good in that regards through his career.

Scheifele I've felt has been overrated in his 2way game here, BUT I have been impressed none the less. He's a rookie so I expect these things except with super-elite prospects. He's also improved through the season here and if you isolate only mins with Kane the numbers are promising.

 

[garret9]

Quick question, I remember reading this somewhere, but it escapes me. Do you know who calculated this, and what the other components/percentages are? A link would be much appreciated also if you have one.

So I did some research in this and it's really hard to find because it was never one article...

It's a combination of multiple articles between Awad, Likens, Ferrari and Desjardins and their subsequent conversations between the authors in their comments sections.

The big quote is here:

Together, Fenwick/Corsi and Luck account for around 3/4 of team winning percentage. What's the remainder? Goaltending talent - which Tom Awad estimates at about 5% - and special teams, along with a very small sliver that's due to shooting talent and the oft-mentioned "shot quality."

but the rest is muddied around everywhere.

From all the articles that I've pulled together lately I got the breakdown like this:
* 70-80% of winning percentage is accounted for by Corsi/Fenwick + luck
* Luck has been estimated to be around 20-30% of that ^ so Corsi/Fenwick would be 40-60% then
* remainder is mix of everything else, with ST being the major chunk
* Awad estimated goaltending around 5%
* Desjardins estimated shot quality (combined for and against) also around 5%