Speculation: Jets General Rumour, Trade, Free Agent and Waiver Speculation 16-17 Part XVI

[BigZ65]

I just can't see betting that heavily on Myers. If anything I see him being traded wholesale at the deadline in terms of moves involving the D. Maybe you get a decent 4-5 LHD and another asset. Don't have to worry about losing anyone from your third line if that D is young enough.

 

[Mortimer Snerd]

Ah thanks!

Ya probably somewhere in the middle but slightly closer to Stempniak value than Tlusty.



Ya but his TOI and 5v4 TOI is way down and most GMs (not all) will get that.

To your last question: *guy-who-wrote-the-Jets-should-sign-Stempniak-as-bottom-six-depth-in-the-summer-before-the-Jets-traded-for-Stempniak raises his hand.*

Ya, I think there might have been 3-4 guys who could raise their hands.

 

[Mortimer Snerd]

I just can't see betting that heavily on Myers. If anything I see him being traded wholesale at the deadline in terms of moves involving the D. Maybe you get a decent 4-5 LHD and another asset. Don't have to worry about losing anyone from your third line if that D is young enough.

Trouble is Myers' injury. He has to play for a while to show that he is not broken before we get anything like value back.

 

[garret9]

Ya, I think there might have been 3-4 guys who could raise their hands.

I wrote this the summer before:

Lee Stempniak - Age: 31 - Pos: RW - Last CapHit: 2.75 mill - 73 GP, 12 G, 22 A, 34 PTS

TOI/60 S/60 Penalty +/- Corsi% relCorsi% O/D ZS% rel O/D ZS% PDO
18.1 7.9 -1 47.5% +1.9% 47.4% -0.9% 96.9

Stempniak offers another alternative at right wing. His PDO plummeted to very low levels this past season, meaning this is a perfect time to buy low on him. While not a top scorer, Stempniak's offensive abilities are about 10 points better than this season, but was loss due to unlucky percentages. He also adds value in being able to take both penalty kill and power play minutes when needed. His two way numbers do place some concerns, although not when looking relative to the other players on his team, but a much larger offensive zone start usage may take care of that.

 

[csk]

I also like Slavin and Pesce on Carolina.

Blockbuster (assuming Trouba won't sign long-term):

To Carolina: Trouba and Perreault

To Winnipeg: Slavin and Aho

Neither Slavin nor Aho need to be protected in the expansion draft.

Jets protect:

Forwards
Scheifele
Little
Wheeler
Lowry
Dano
Armia
Copp

Defense
Buff
Enstrom
Myers

Trade Connor in a package for a good young RHD who can play top-4.

2017/18 Roster:

Ehlers-Scheifele-Laine
Aho-Little-Wheeler
Dano-Lowry-Armia
Copp-Petan-Roslovic

Slavin-Buff
Morrissey-xRD
Enstrom-Myers

Postma
Poolman
Chiarot

This looks worse than our current team somehow

 

[Mortimer Snerd]

I just can't see betting that heavily on Myers. If anything I see him being traded wholesale at the deadline in terms of moves involving the D. Maybe you get a decent 4-5 LHD and another asset. Don't have to worry about losing anyone from your third line if that D is young enough.

Trouble is Myers' injury. He has to play for a while to show that he is not broken before we get anything like value back.

 

[Mortimer Snerd]

I wrote this the summer before:

Didn't mean to suggest you weren't a hand raiser. The others were probably just following your lead.

 

[JetsFan815]

Didn't mean to suggest you weren't a hand raiser. The others were probably just following your lead.

Not me. I was pro-Stempniak, long before being pro-Stempniak was cool around these parts.

 

[YWGinYYZ]

Not me. I was pro-Stempniak, long before being pro-Stempniak was cool around these parts.

There were many that liked Stempniak, and wanted him re-signed. Me for instance. It's not as rare as you're making it out to be, you just happen to be the most vocal about it.

 

[garret9]

Dawson updated XPM so now I have a more accurate estimate of WAR.*

Here is average depth position as depicted by performance in WAR and Game Score:

I'm estimating WAR using a weighted combination of XPM and scoring. XPM and player scoring data as of this morning (so not including today's game).

Game Score was as of two days ago (so not including Pittsburgh game).

* The XPM data was a few weeks old. In addition, we made some slight changes that has increased both the predictive level and the reputability to the metric.

 

[garret9]

Dawson updated XPM so now I have a more accurate estimate of WAR.*

Here is average depth position as depicted by performance in WAR and Game Score:

I'm estimating WAR using a weighted combination of XPM and scoring. XPM and player scoring data as of this morning (so not including today's game).

Game Score was as of two days ago (so not including Pittsburgh game).

* The XPM data was a few weeks old. In addition, we made some slight changes that has increased both the predictive level and the reputability to the metric.

Some movers and shakers from last update...

Forwards
1) Scheifele has fallen a bit from a bonafide first liner to fringe first liner
2) Laine has moved up from bonafide second liner to fringe first liner
3) Lowry has fallen from fringe second liner to bonafide 3rd liner
4) Dano had an error in his data and has fallen, but is still good
5) Thorburn is still not a NHLer but by a lesser degree
6) Copp is being pulled up by his scoring a bit more than before

Defense
1) Byfuglien has moved up to bonafide from fringe #1D, but is still not like Trouba (who is elite)
2) Morrissey, as expected, has been moving up into bonafide top 4 Dman
3) Postma, not as expected, has moved up but like Copp has some unsustainable scoring going on
4) Stuart and Chiarot are still bad

 

[Daximus]

Dawson updated XPM so now I have a more accurate estimate of WAR.*

Here is average depth position as depicted by performance in WAR and Game Score:

I'm estimating WAR using a weighted combination of XPM and scoring. XPM and player scoring data as of this morning (so not including today's game).

Game Score was as of two days ago (so not including Pittsburgh game).

* The XPM data was a few weeks old. In addition, we made some slight changes that has increased both the predictive level and the reputability to the metric.

Is there any way to get charts like these for every team? Or are you making them?

 

[garret9]

Is there any way to get charts like these for every team? Or are you making them?

I'm making them myself and there's a lot of manual work to it.

Process is follows:
1) XPM is available through DTMAH and he updates about once every 2 weeks.
2) I take a player's 5v5 goals, primary assists, and secondary assists from Corsica.
3) I combine the goals, primary assists, and secondary assists in a non-1:1:1 weighting, which is different for forwards vs defenders.
4) XPM accounts for 54% of a defender's offense value and 100% of their defensive value. Forwards are the same but XPM is 52% instead of 54% as scoring is relatively more improtant for forwards.
5) The adjusted scoring really only accounts for 60% of BPM (XPM's counterpart) but I pretend like it accounts for everything.
6) This gives us an estimate of WAR that accounts for about 90% of WAR's actual inputs.
6) Game Score is just provided through Dom and he updates about the same often as DTMAH does.
7) I use their ranking in WAR to estimate what line their performance is in. GS already says whether they are performing to a certain line.
8) I average the two (WAR and GS). So if someone is a 1st liner in WAR and 2nd liner in GS, they get a 1.5.
9) I then build the chart in excel.

 

[garret9]

If you want access to Game Score: https://gamescorecharts.wordpress.com/2016/12/27/depth-charts/

That gives you 50% of the above value.

Note: The largest thing you will notice is defensive players are not valued highly. Game Score's largest weakness is catching the impact of defense.

 

[Daximus]

I'm making them myself and there's a lot of manual work to it.

Process is follows:
1) XPM is available through DTMAH and he updates about once every 2 weeks.
2) I take a player's 5v5 goals, primary assists, and secondary assists from Corsica.
3) I combine the goals, primary assists, and secondary assists in a non-1:1:1 weighting, which is different for forwards vs defenders.
4) XPM accounts for 54% of a defender's offense value and 100% of their defensive value. Forwards are the same but XPM is 52% instead of 54% as scoring is relatively more improtant for forwards.
5) The adjusted scoring really only accounts for 60% of BPM (XPM's counterpart) but I pretend like it accounts for everything.
6) This gives us an estimate of WAR that accounts for about 90% of WAR's actual inputs.
6) Game Score is just provided through Dom and he updates about the same often as DTMAH does.
7) I use their ranking in WAR to estimate what line their performance is in. GS already says whether they are performing to a certain line.
8) I average the two (WAR and GS). So if someone is a 1st liner in WAR and 2nd liner in GS, they get a 1.5.
9) I then build the chart in excel.

The one thing I've always thought was weird is when people use the 1 through 6 usage definition for dmen rather than the 1, 2, 3 pairings and 4 for depth to explain usage.

We don't say 1 through 12 for forwards, I don't understand why we say 1 through 6 for dmen. I guess it made more sense when teams placed less emphasis on handedness and just paired guys willy nilly but now that handedness is pretty much widespread you'd think we would just use the 1, 2 and 3 for pairings and 4 for depth. Dmen pretty much now have a sub position like forwards that is dependent on their handedness. So when we say a 1LW we mean a 1st line LW. When we say a 1LD we mean a 1st pairing left handed dman. I think it's time we just go to pairings.

Sorry it's just something that has always bothered me.

 

[garret9]

The one thing I've always thought was weird is when people use the 1 through 6 usage definition for dmen rather than the 1, 2, 3 pairings and 4 for depth to explain usage.

We don't say 1 through 12 for forwards, I don't understand why we say 1 through 6 for dmen. I guess it made more sense when teams placed less emphasis on handedness and just paired guys willy nilly but now that handedness is pretty much widespread you'd think we would just use the 1, 2 and 3 for pairings and 4 for depth. Dmen pretty much now have a sub position like forwards that is dependent on their handedness. So when we say a 1LW we mean a 1st line LW. When we say a 1LD we mean a 1st pairing left handed dman. I think it's time we just go to pairings.

Sorry it's just something that has always bothered me.

The reason I do has more to do with impact.

The gap in impact between with improving a #1->#4 defender is fairly large. A lot larger than #1->#6 forward.

 

[CaptainChef]

Updated some numbers. Then I combined (averaged) player depth positions based on performance of DTM's WAR model and Dom's Game Score model.

Gives a good shot of where the Jets do well/poorly and need help:

when you apply these numbers to who we need to protect in the expansion draft, seems clear that the Jets aren't going to want to give up Lowry. And not likely MP either. Dano at 2 is another keeper.

Getting more & more obvious we are going to have to find a way to get down to 3 defensemen to protect.

Edit: see the numbers have changed a fair bit, but doesn't really change the bottom line; we need to find a way to do 7+3+1

 

[Daximus]

The reason I do has more to do with impact.

The gap in impact between with improving a #1->#4 defender is fairly large. A lot larger than #1->#6 forward.

Wouldn't the .5 decimals account for that though?

 

[garret9]

Wouldn't the .5 decimals account for that though?

Not exactly. The decimal numbers just mean there is a disagreement between the two models.

A 1.5 is saying one model is saying 1 while another is saying 2.

 

[JetsFan815]

There were many that liked Stempniak, and wanted him re-signed. Me for instance. It's not as rare as you're making it out to be, you just happen to be the most vocal about it.

I've never claimed it was rare. Although many were not as outraged about it when it happened as I would have liked. There was a lot of "he's a replaceable dime-a-dozen bottom-6 player" attitude. Most of the outrage started after he started tearing it up with the Devils.

 

[JetsFan815]

I'm making them myself and there's a lot of manual work to it.

Process is follows:
1) XPM is available through DTMAH and he updates about once every 2 weeks.
2) I take a player's 5v5 goals, primary assists, and secondary assists from Corsica.
3) I combine the goals, primary assists, and secondary assists in a non-1:1:1 weighting, which is different for forwards vs defenders.
4) XPM accounts for 54% of a defender's offense value and 100% of their defensive value. Forwards are the same but XPM is 52% instead of 54% as scoring is relatively more improtant for forwards.
5) The adjusted scoring really only accounts for 60% of BPM (XPM's counterpart) but I pretend like it accounts for everything.
6) This gives us an estimate of WAR that accounts for about 90% of WAR's actual inputs.
6) Game Score is just provided through Dom and he updates about the same often as DTMAH does.
7) I use their ranking in WAR to estimate what line their performance is in. GS already says whether they are performing to a certain line.
8) I average the two (WAR and GS). So if someone is a 1st liner in WAR and 2nd liner in GS, they get a 1.5.
9) I then build the chart in excel.

Sounds like something that should be automated, no human should have to do it more than once

 

[Daximus]

Not exactly. The decimal numbers just mean there is a disagreement between the two models.

A 1.5 is saying one model is saying 1 while another is saying 2.

So essentially 1-2.5 is a first pairing, 3-4.5 is a 2nd pairing and 5-6.5 is a 3rd and depth would be 7-8 and anything after is pure garbage.

 

[Daximus]

I've never claimed it was rare. Although many were not as outraged about it when it happened as I would have liked. There was a lot of "he's a replaceable dime-a-dozen bottom-6 player" attitude. Most of the outrage started after he started tearing it up with the Devils.

I think the majority of posters knew even without the scoring that he was a workhorse even in the bottom 6. The first few games with us I noticed that. Guy had a motor that just never quit which basically explained our identity that entire season. I see a little bit of that play style in Matthias. Though I think Stemp was more effective. Sometimes I imagine if we would have kept both Stemp and Frolik.

Ehlers - Scheif - Wheeler
Perro - Little - Laine
Stemp - Lowry - Frolik
Matthias/Dano - Copp - Armia/Dano
Dano

That is a bottom 6 that could give opposing teams nightmares. The top 6 already has that ability. Then you imagine Jack and Kyle knocking on the door next season.

 

[Hobble]

If we kept Frolik and Stempniak over Stafford, we probably wouldn't have stunk hard enough to win Laine.

 

[Daximus]

If we kept Frolik and Stempniak over Stafford, we probably wouldn't have stunk hard enough to win Laine.

See and that is the paradox. This might just be the brightest timeline.