Line Combos: 2018-19 Starting Roster II

[Maukkis]

If buff is out a month then what's our d lines look like?

Horrific, I presume.

Mo-Trouba
Chiarot-Myers
Morrow-Kulikov (thinking it was Kulikov who has played the right side when need has arisen in the past)
Niku (for some reason...)

If 44 and 8 are separated, I reckon Chiarot plays with Trouba and Morrissey drops down a pair to play with Myers. That pairing has been a travesty in the past...

 

[Psych0dad]

^TLDR version:

The best practice with looking at where Laine plays best is still with looking at Corsi and ignoring GF/GA due to small samples.

Or simply:

Laine's 45% Corsi > than an average player's 45% Corsi, but Laine's 50% Corsi > than Laine's 45% Corsi.

Agreed with most of it. Do not agree with small sample size. And the natural fix to small sample size is getting a bigger one.

 

[garret9]

Agreed with most of it. Do not agree with small sample size. And the natural fix to small sample size is getting a bigger one.

As Whileee and I have discussed previously in the stats thread, sample sizing and confidence can be measured.

I mean no offense, but your agreement is irrelevant to what has been proven.

It would take about 3-4 years of Laine being with the same linemates (including d-men) without any line swapping for GF/GA to equate to same confidence level you get from one season's worth of Corsi with his most common linemates.

There's also two fixes to sample sizing issues...
1) larger sample - with the potential of wasting time in suboptimal situations
2) find a measure that reaches higher confidence levels sooner - ie: use Corsi

 

[Psych0dad]

None taken, that's how things work.

So corsi is a better predictor of future GF60-GA60, than past GF60-GA60?

Because the aim should be to increase the positive difference in GF60-GA60, right?

So does corsi predict that better?

 

[garret9]

None taken, that's how things work.

So corsi is a better predictor of future GF60-GA60, than past GF60-GA60?

Because the aim should be to increase the positive difference in GF60-GA60, right?

So does corsi predict that better?

Yes:
Remembering Dellow: A few graphs to convince you on Corsi

 

[EXTRAS]

Horrific, I presume.

Mo-Trouba
Chiarot-Myers
Morrow-Kulikov (thinking it was Kulikov who has played the right side when need has arisen in the past)
Niku (for some reason...)

If 44 and 8 are separated, I reckon Chiarot plays with Trouba and Morrissey drops down a pair to play with Myers. That pairing has been a travesty in the past...

You think niku is bad?

 

[Maukkis]

You think niku is bad?

Nope. He won't make the roster, though, which is a shame. We need a puck mover with Buff out.

 

[Psych0dad]

Yes:
Remembering Dellow: A few graphs to convince you on Corsi

So in short, we shouldn't aim for most positive GF60-GA60 difference, we should aim for best corsi?

Wagon before the horse?

 

[Psych0dad]

Did you have an opinion on optimized lines?

 

[garret9]

So in short, we shouldn't aim for most positive GF60-GA60 difference, we should aim for best corsi?

Wagon before the horse?

I would change it to: aiming for the best GF60-GA60 performance you would look at past Corsi performance.

That's putting the horse before the wagon.

 

[jepjepjoo]

I would change it to: aiming for the best GF60-GA60 performance you would look at past Corsi performance.

That's putting the horse before the wagon.

16-17Top5 CF%
1.Wheeler
2.Perreault
3.Myers
4.Morrisey
5.Little
vs
17-18Top5 GF%
1.Roslovic
2.Tanev
3.Laine
4.Enstrom
5.Copp
....
10.Wheeler
11.Perreault
...
15.Morrissey
...
17.Myers
...
19.Little

---------------------
16-17Top5 CF%
1.Wheeler
2.Perreault
3.Myers
4.Morrisey
5.Little
vs
17-18Top5 GF/60
1.Laine
2.Trouba
3.Scheifele
4.Ehlers
5.Wheeler
...
7.Morrissey
...
13.Perreault
14.Myers
...
20.Little
---------------------
16-17Top5 GF%
1.Myers
2.Dano
3.Copp
4.Scheifele
5.Laine
vs
17-18Top5 GF%
1.Roslovic
2.Tanev
3.Laine
4.Enstrom
5.Copp
6.Scheifele
...
17.Myers

 

[garret9]

16-17Top5 CF%
1.Wheeler
2.Perreault
3.Myers
4.Morrisey
5.Little
vs
17-18Top5 GF%
1.Roslovic
2.Tanev
3.Laine
4.Enstrom
5.Copp
....
10.Wheeler
11.Perreault
...
15.Morrissey
...
17.Myers
...
19.Little

---------------------
16-17Top5 CF%
1.Wheeler
2.Perreault
3.Myers
4.Morrisey
5.Little
vs
17-18Top5 GF/60
1.Laine
2.Trouba
3.Scheifele
4.Ehlers
5.Wheeler
...
7.Morrissey
...
13.Perreault
14.Myers
...
20.Little
---------------------
16-17Top5 GF%
1.Myers
2.Dano
3.Copp
4.Scheifele
5.Laine
vs
17-18Top5 GF%
1.Roslovic
2.Tanev
3.Laine
4.Enstrom
5.Copp
6.Scheifele
...
17.Myers

And what happens when you surpass the sample beyond just a few players... Oh...

 

[jepjepjoo]

And what happens when you surpass the sample beyond just a few players... Oh...

I still think that people put too much weight on a simple stat such as corsi when the game of hockey is so complex. In Baseball for example you can pretty much tell the worth of a player just by looking at his stats. Hockey has just too many variables.

 

[Psych0dad]

16-17Top5 CF%
1.Wheeler
2.Perreault
3.Myers
4.Morrisey
5.Little
vs
17-18Top5 GF%
1.Roslovic
2.Tanev
3.Laine
4.Enstrom
5.Copp
....
10.Wheeler
11.Perreault
...
15.Morrissey
...
17.Myers
...
19.Little

---------------------
16-17Top5 CF%
1.Wheeler
2.Perreault
3.Myers
4.Morrisey
5.Little
vs
17-18Top5 GF/60
1.Laine
2.Trouba
3.Scheifele
4.Ehlers
5.Wheeler
...
7.Morrissey
...
13.Perreault
14.Myers
...
20.Little
---------------------
16-17Top5 GF%
1.Myers
2.Dano
3.Copp
4.Scheifele
5.Laine
vs
17-18Top5 GF%
1.Roslovic
2.Tanev
3.Laine
4.Enstrom
5.Copp
6.Scheifele
...
17.Myers

Thank you, I had a feeling that's what it would show.

And I bet GF60-GA60 differential between seasons was a better predictor between the seasons too. Closer to real results.

 

[voyageur]

Well there should be a change coming now.

I'd like to see

Laine-Scheif-Wheeler because the kid needs more shots, and those are two guys who will create chances.

Perreault-Little-Ehlers would bring a lot of pace.

Connor-Roslovic-Appleton would keep the pace high

Tanev-Lowry-Copp will get the puck in the right zone.

Defense

Morrissey-Trouba more minutes, 25 at least.

Kulikov-Niku. Kulikov is playing very stable. Stable is what Niku needs to be at his best. They could push the pace.

Chiarot-Myers. They have 3 rounds of playoff experience. Going to be your PK2 now. Could have some rough shifts but Myers can add some offense with Buff out.

Morrissey QBing the 1st PP. Maybe try Niku and Trouba on the 2nd.

 

[talitintti]

The model of hockey at EV is still the same:
* Shot quantity (try to make more shots and allow fewer)
* Shot quality (try to push your shots closer and to a less wide angle)
* Finishing talent / goalie talent (maximize on chances and minimize opponent chances)

Laine's finishing talent doesn't work the same way relative to league average in high quality chances, his talent comes from stretching the high quality shooting area. So, in a way, that breaks the model.

 

[garret9]

Laine's finishing talent doesn't work the same way relative to league average in high quality chances, his talent comes from stretching the high quality shooting area. So, in a way, that breaks the model.

That literately doesn’t break the model and is exactly what I was saying: shift.

In fact, there was an xG model that uses sh% regression factor to account for finishing talent. All it does is add a multiplier. Model not broken.

 

[JetsFan815]

Maurice with the super blender... Jets have been really sluggish and he's trying to find some life.

I must have blinked and missed the super blender because it was all the same adjustments to me (Ehlers/Connor swapping) other than a few shifts for Perreault on the 2nd line from mid-2nd to mid-3rd.

The aversion to using Lowry in anything other than a checking role, aversion to splitting up Scheifele-Wheeler, aversion to putting Perreault on one of the top-2 lines, aversion to using forwards 10-12 in any reasonable capacity, handcuff the Jets from a "line blender". When all those (self-imposed) constraints are taken into account, the only adjustment or blending left to do is Connor-Ehlers swap, that's pretty much all we do when things are not working.

The 2nd line has been twisting in the winds for months and has some of the worst numbers in the entire league (no exaggeration) yet I don't see anything being tried to change the course with them other than just retrying ELL and CLL. I don't know how you can be a serious contender in this league with your 2nd line being one of the worst lines in the league. There is time to fix it before the playoffs but I don't see any experimentation to rescue that line.

 

[garret9]

I still think that people put too much weight on a simple stat such as corsi when the game of hockey is so complex. In Baseball for example you can pretty much tell the worth of a player just by looking at his stats. Hockey has just too many variables.

Thank you, I had a feeling that's what it would show.

And I bet GF60-GA60 differential between seasons was a better predictor between the seasons too. Closer to real results.

For fun, I repeated same exercise with JUST Jets players, from 2011 to present.

Two things:

1) Same relationship in how the gap between CF and GF shrinks.

Binning players by how much they differed in Corsi% and Goal% and then looking at how much they differed in year two:

CF differed GF by >10 points: average 15.9, next year 1.3
CF differed GF by 5 to 10 points: average 7.2, next year 0.1
CF differed GF by 0 to 5 points: average 2.7, next year -0.5
CF differed GF by 0 to -5 points: average -2.1, next year 1.3
CF differed GF by -5 to -10 points: average -7.1, next year 0.5
CF differed GF by <-10 points: average -24.1, next year -1.8

2) Corsi predicted next year performance better.

Corsi in year one had over twice the relationship (R^2) as Goals in year one for Goals in year two for a player. (I used % form so CF / (CF + CA) and GF / (GF + GA) instead of CF - CA since 24-22 is closer to even than 4-2).

 

[KingBogo]

I must have blinked and missed the super blender because it was all the same adjustments to me (Ehlers/Connor swapping) other than a few shifts for Perreault on the 2nd line from mid-2nd to mid-3rd.

The aversion to using Lowry in anything other than a checking role, aversion to splitting up Scheifele-Wheeler, aversion to putting Perreault on one of the top-2 lines, aversion to using forwards 10-12 in any reasonable capacity, handcuff the Jets from a "line blender". When all those (self-imposed) constraints are taken into account, the only adjustment or blending left to do is Connor-Ehlers swap, that's pretty much all we do when things are not working.

The 2nd line has been twisting in the winds for months and has some of the worst numbers in the entire league (no exaggeration) yet I don't see anything being tried to change the course with them other than just retrying ELL and CLL. I don't know how you can be a serious contender in this league with your 2nd line being one of the worst lines in the league. There is time to fix it before the playoffs but I don't see any experimentation to rescue that line.

What experiment is needed to fix the 2nd line? You either move Laine up with Scheifele or you don't. As you indicated Connor and Ehlers are moved freely back and forth on the top 2 lines with some mixed results. They both add something different to the 1st line but can't carry the 2nd line. Little's numbers would suggest an average #2 center at this point in his career. It has become evident that this is not enough to carry Laine with either of Ehlers or Connor.

So the current option is keeping Little and Laine together to see if Laine can learn to help carry a line. IMO this is the desired outcome and why Maurice is so persistent on seeing if Laine can take the next step. It is also likely that Maurice has Chevy's support with this as he needs to figure out how much is he willing to spend on Laine's upcoming contract. Since the Jets are a strong contender and well on the way to securing a playoff spot there is no need to panic.

If the Little-Laine combo is still struggling as we approach the TD, maybe we see the Maurice give up this season the hope that Laine can drive a line and puts him with Scheifele or he dips into the future and picks up a #2 center like he did at last year's TD. Though if that is the case I'd prefer to really dig deep and get someone younger and with a bit of term.

IMO the early hope that Rosie could take the next couple steps and surpass Little as the #2 is likely gone by the way side for this season at least.

 

[Psych0dad]

So the current option is keeping Little and Laine together to see if Laine can learn to help carry a line. IMO this is the desired outcome and why Maurice is so persistent on seeing if Laine can take the next step. It is also likely that Maurice has Chevy's support with this as he needs to figure out how much is he willing to spend on Laine's upcoming contract. Since the Jets are a strong contender and well on the way to securing a playoff spot there is no need to panic.

I am 100% sure that Laine would grow into a "line driver" without this silly forced assignment, and probably faster than he will this way. You can learn to be a line driver while doing what you do best. I think it would be in Jets best interest to play him to his strengths and not to his weaknesses at this point in development. Confidence and success grows a player faster than suffering and struggling. It's an old, unproven and stale mentality that you have to suffer to succeed. Stick, no carrot.

Jets would get much more out of him immediately, not in 4 years. And he would still learn (due to physical development and experience) how to become even more of a complete player.

There are better ways to reach that desired outcome. Who ever drew this plan on the board is a moron.

I do think that he has Chevy's blessing in limiting Laine. And that is a long term savings effort. And I think it's a doomed attempt that will just burn the bridge they wanted to build. It's the wrong approach and could end up costing them their franchise player. And they will have only themselves to blame.

 

[surixon]

What experiment is needed to fix the 2nd line? You either move Laine up with Scheifele or you don't. As you indicated Connor and Ehlers are moved freely back and forth on the top 2 lines with some mixed results. They both add something different to the 1st line but can't carry the 2nd line. Little's numbers would suggest an average #2 center at this point in his career. It has become evident that this is not enough to carry Laine with either of Ehlers or Connor.

So the current option is keeping Little and Laine together to see if Laine can learn to help carry a line. IMO this is the desired outcome and why Maurice is so persistent on seeing if Laine can take the next step. It is also likely that Maurice has Chevy's support with this as he needs to figure out how much is he willing to spend on Laine's upcoming contract. Since the Jets are a strong contender and well on the way to securing a playoff spot there is no need to panic.

If the Little-Laine combo is still struggling as we approach the TD, maybe we see the Maurice give up this season the hope that Laine can drive a line and puts him with Scheifele or he dips into the future and picks up a #2 center like he did at last year's TD. Though if that is the case I'd prefer to really dig deep and get someone younger and with a bit of term.

IMO the early hope that Rosie could take the next couple steps and surpass Little as the #2 is likely gone by the way side for this season at least.

There are plenty of permutations that could be tried. We aren't limit3d to just swapping a couple of players.

We had a great second line in Perrault Little Ehlers last year that scored really well and also drove possession.

Conner/Laine Scheiele Wheeler
Perrault -Little Ehlers
Conner/Laine- Petan - Roslovic (Sheltered offensive role)
Copp - Lowry - Tanev defensive role

I don't see the issue in switching things up and giving some of the kids a chance. As mentioned in another thread our second line is anemic and our third and fourth lines are struggling to produce, now is the time to experiment and take some chances to see what you have. We don't need to be rigid in our line combos.

 

[KingBogo]

There are plenty of permutations that could be tried. We aren't limit3d to just swapping a couple of players.

We had a great second line in Perrault Little Ehlers last year that scored really well and also drove possession.

Conner/Laine Scheiele Wheeler
Perrault -Little Ehlers
Conner/Laine- Petan - Roslovic (Sheltered offensive role)
Copp - Lowry - Tanev defensive role

I don't see the issue in switching things up and giving some of the kids a chance. As mentioned in another thread our second line is anemic and our third and fourth lines are struggling to produce, now is the time to experiment and take some chances to see what you have. We don't need to be rigid in our line combos.

I think the biggest difference in our opinions is you see more in our 4th line players than I do. I've all but given up on Petan ever being an everyday player, just too many holes in his game. I agree with you on Lemieux, but I'd put him in the same not ready for prime time pile, just with less talent. I'm still hopeful for Appleton, but he likely never rises much beyond what we see now. Starting to doubt Rosie will ever be much more than a Copp level player (3rd line winger, 4th line center). So to spread out as you suggest puts Connor or Laine in a difficult spot to produce. IMO we have as good as top 6 as there is in the league and added to this are 3 good forwards in my descending order Lowry, MP, Tanev. After that nothing left over amounts to anything special IMO. Components of an adequate 4th but that is about it. I'd be surprised if we didn't see Chevy pick up a couple veterans to take on this role for the playoffs.

 

[jepjepjoo]

For fun, I repeated same exercise with JUST Jets players, from 2011 to present.

Two things:

1) Same relationship in how the gap between CF and GF shrinks.

Binning players by how much they differed in Corsi% and Goal% and then looking at how much they differed in year two:

CF differed GF by >10 points: average 15.9, next year 1.3
CF differed GF by 5 to 10 points: average 7.2, next year 0.1
CF differed GF by 0 to 5 points: average 2.7, next year -0.5
CF differed GF by 0 to -5 points: average -2.1, next year 1.3
CF differed GF by -5 to -10 points: average -7.1, next year 0.5
CF differed GF by <-10 points: average -24.1, next year -1.8

2) Corsi predicted next year performance better.

Corsi in year one had over twice the relationship (R^2) as Goals in year one for Goals in year two for a player. (I used % form so CF / (CF + CA) and GF / (GF + GA) instead of CF - CA since 24-22 is closer to even than 4-2).

last 2 and a half seasons combined GF% - CF% (forwards) :

McDavid +6.85
Kucherov +4.9
Ovechkin +7.7
Scheifele +6.29
Zucker +10.19
Kuznetsov +8.93
Marner 4.96
Point +5.54
Radulov +5.25
Laine +5.26
Matthews +7.91
Kessel +4.94
Backstrom +11.2
Granlund +8.67
Benn +6.78
Karlsson +8.56
Atkinson +7.98
Rakell +5.15
Forsberg +4.48
Oshie +12.19
Smith +5.21
Anderson +8.67
Jarnrok +6.78
Wilson +7.48
Point +5.54
Wennberg +5.25
Brown +4.48

Larsson -12.12
Crouse -13.04
McKenzie -10.75
Richardson -12.33
J.T Brown -12.17
Wingels -10.42
Ma. Granlund -10.38
Nordstrom -11.41
McGinn -13.25
Martinook -9.95
Z.Smith -6.94
Abdelkader -6.07
Girgesons -7.19
Comeau -7.67
Berglund -9.97
Schaller -7.84
Lehtera -7.34
Bennett -9.13

Seems like good players beat the average and bad players don't... go figure.

 

[10Ducky10]

This would be my lineup for tonight's game...

KC Scheif Appleton
MP Little Wheeler
Ehlers Roslo Laine
Copper Lowry Tanev

JMo Trouba
Kuli Myers
Chiarot Morrow

Brossoit Helle