2015-2016 Statistical And Underlying Metrics Thread

[Daximus]

Since deadline:

Where is Gump?

Burmi really has upped his game since Ladds departure and him getting more ice time. Petan has looked much better now than he did during his first stint. Seems that time in the AHL really helped him out.

 

[Aavco Cup]

 

[garret9]

Friend did this, it's the expected Goal% of Jets/Thrashers with each tick being a different season and colour being a different coach:

 

[Evil Little]

Friend did this, it's the expected Goal% of Jets/Thrashers with each tick being a different season and colour being a different coach:

Everything after the blue is just another way of telling the story of Ondrej Pavelec, right?

 

[Daximus]

Friend did this, it's the expected Goal% of Jets/Thrashers with each tick being a different season and colour being a different coach:

Anyway to see other teams graphs over the same or similar period of time?

 

[Gm0ney]

Everything after the blue is just another way of telling the story of Ondrej Pavelec, right?

I don't think the goalie would impact the "Expected Goals %".

 

[garret9]

Anyway to see other teams graphs over the same or similar period of time?

He hasn't released it yet.

I'd link it when he does, but it likely be HG which I can't link to as I have a vested interest in the site (although no financial as it generates no income) and I only am a partner through Jets Nation, not Hockey Graphs.

So, keep your eye out for a ctompkins article on HG in the next few weeks and you can link it.

I don't think the goalie would impact the "Expected Goals %".

Yes but they'd impact the difference between observed and expected goal%, which I think is his point.

 

[Gm0ney]

Yes but they'd impact the difference between observed and expected goal%, which I think is his point.

Is that what's it's showing? Label your graphs so we don't have to reverse engineer or guess at everything!

 

[White Out 403*]

You mean to tell me Stuart and Pavelec are huge anchors on this team?

What?

It's honestly less fun being this right all the time. All the surprise and unknown of a lot of things are taken away.

 

[csk]

Is that what's it's showing? Label your graphs so we don't have to reverse engineer or guess at everything!

He did label it, it says expected goal diff in the comment

 

[garret9]

Is that what's it's showing? Label your graphs so we don't have to reverse engineer or guess at everything!

Not my graph, so I can't label it, but I did indicate just above it:

it's the expected Goal% of Jets/Thrashers with each tick being a different season and colour being a different coach

 

[Evil Little]

So...

How does one change the sample seasons in the OwnthePuck charts...?

 

[garret9]

So...

How does one change the sample seasons in the OwnthePuck charts...?

Can't yet.

Dom still has work to do with the programming since he now pulls from Corsica instead of HockeyAnalysis.

 

[csk]

Can't yet.

Dom still has work to do with the programming since he now pulls from Corsica instead of HockeyAnalysis.

I was wondering why that got changed, good to know

 

[garret9]

I was wondering why that got changed, good to know

He went to Corsica because the data adjusts for score and venue unlike HockeyAnalysis, and also because he could change the goals to expected goal model (which differs from the DTMAH/Hockey-Graphs model slightly).

 

[garret9]

Playoff success has so much luck involved.

 

[puck stoppa]

Im thinking ANA STL in the west, TB WSH in the east for the final four.

 

[garret9]

Vancouver Hockey Analytics Conference starts tomorrow. It will be streamed.
I'll be speaking ~10:15 PT on goaltending and the hot hand.

Spoiler: found no evidence of playing the hot hand is a good thing, especially with Pavelec.

 

[powder88]

Any update of the Jets penalty kill, now that Stuart is back? My sense was that it had improved during his departure and has bottomed out again with his return. What say you, numbers gods?

 

[garret9]

now the second worst special teams team, not the worst:

 

[powder88]

now the second worst special teams team, not the worst:

Thanks Garret, but I was curious about whether we can tease out the impact of Stuart In versus Stuart Out. I think I was able to come close here:

Stuart missed 18 games from Feb 16 to March 22, inclusive. During that time, the Jets gave up 11 PPG on 57 times shorthanded, for an 80.7% PK

Through the first chunk of the season the Jets gave up 45 goals on 203 chance, for a 77.8% PK

In the most recent chunk, with Stuart back on the PK, there have been 5 GA on 22 attempts, or 77.3%

In total then, the Jets killed penalties at a 77.8% clip when Stuart played a big role on the PK and 80.7% without.

If they had been able to kill penalties at that clip all season, they would have given up about 7 fewer goals on the PK.

 

[garret9]

Thanks Garret, but I was curious about whether we can tease out the impact of Stuart In versus Stuart Out. I think I was able to come close here:

Stuart missed 18 games from Feb 16 to March 22, inclusive. During that time, the Jets gave up 11 PPG on 57 times shorthanded, for an 80.7% PK

Through the first chunk of the season the Jets gave up 45 goals on 203 chance, for a 77.8% PK

In the most recent chunk, with Stuart back on the PK, there have been 5 GA on 22 attempts, or 77.3%

In total then, the Jets killed penalties at a 77.8% clip when Stuart played a big role on the PK and 80.7% without.

If they had been able to kill penalties at that clip all season, they would have given up about 7 fewer goals on the PK.

7 fewer goals *on average* equates to about 1 and a bit wins.

 

[powder88]

7 fewer goals *on average* equates to about 1 and a bit wins.

Great...now can we talk about the extra GA caused by him being on the ice for the 5-on-5 play???

 

[Aavco Cup]

I wonder if Garret is the analytics guy helping Lawless?

 

[ps241]

I wonder if Garret is the analytics guy helping Lawless?

I appreciate that Gary actually takes his responsibility seriously because some of the writers (voters) seem to mail it in.