Line Combos: 2018-19 Starting Roster II

[Maukkis]

Tanev-Lowry-Laine was not expected to work, not a bit. It takes a record breaking coach to put Tanev and Laine into the same line.

Connor-Wheeler-Laine actually did fine. Look at their goals for, goals against. They were really effective offensively, meh defensively as expected, but after all net positive with a decent margin.

Sorry, but your claims do not match match reality, goals for vs. goals against. I'm on mobile (as usual) and in a hurry, so I'm not going to get you statistics from natural stat trick, but in short you are just wrong. It's not that those stats weren't repeated here over and over again.

And please, Perreault - Little - Laine was not a really bad line, but still I think they were either net negative or net neutral whereas Laine lines with Copp, Stastny or Scheifele (or Connor & Wheeler) were hugely net positive. Please don't make things up.

All long running Laine lines sans Little have been hugely net positive.

Three letters are all I'm going to sacrifice to debunk your take on CWL.

PDO.

 

[Psych0dad]

Three letters are all I'm going to sacrifice to debunk your take on CWL.

PDO.

Totally irrelevant. GF% and GA% matter. Actual results. The difference between GF/60 and GA/60, how positive it is.

That's what you should focus on, because what matters in winning hockey games is that. Not PDO.

Real results on the scoreboard

VS.

"the sum of a team's shooting percentage and its save percentage."


Are we trying to win games or PDO? Which would you prefer?

For me it's pretty easy decision. I take winning games, over winning pdo or corsi or intangibles or looking good while losing or hustle or perception of supreme effort. What ends up on the board counts. Everything else is noise.

 

[Maukkis]

Totally irrelevant. GF% and GA% matter. Actual results. The difference between GF/60 and GA/60, how positive it is.

That's what you should focus on, because what matters in winning hockey games is that. Not PDO.

Real results on the scoreboard

VS.

"the sum of a team's shooting percentage and its save percentage."


Are we trying to win games or PDO? Which would you prefer?

Try to read the discussion before jumping in. Preferably, try to understand it too, which you clearly did not do.

 

[Psych0dad]

Try to read the discussion before jumping in. Preferably, try to understand it too, which you clearly did not do.

Go ahead and explain then. I think you were suggesting they were a bad line due to PDO after others showed you they did alright in actual results. What did I get wrong?

 

[Romang67]

This isn't the Laine thread.

 

[JetsFan815]

As I have said before, I am in the camp that has no interest in rentals this year. However I would be open to moving assets to get a player with some term on him. With Carolina faltering, Hamilton not being that hot for them and an impatient owner looking for moves to be made I wonder if they would consider:

Dougie Hamilton for Jets 1st, one of Myers/Kulikov, one of Petan/Appleton/Lemieux, next year's 2nd.

A trade like this would give us more flexibility when it comes to moving Trouba next summer and increase the variety of the assets we would be willing to accept in such a trade. Also would really shore up the Jets D.

 

[Whileee]

As I have said before, I am in the camp that has no interest in rentals this year. However I would be open to moving assets to get a player with some term on him. With Carolina faltering, Hamilton not being that hot for them and an impatient owner looking for moves to be made I wonder if they would consider:

Dougie Hamilton for Jets 1st, one of Myers/Kulikov, one of Petan/Appleton/Lemieux, next year's 2nd.

A trade like this would give us more flexibility when it comes to moving Trouba next summer and increase the variety of the assets we would be willing to accept in such a trade. Also would really shore up the Jets D.

Hamilton worries me, mainly in terms of how he'd fit with the team. Maybe that's overblown, but it would give me some pause.

 

[Halberdier]

Three letters are all I'm going to sacrifice to debunk your take on CWL.

PDO.

Oh noes. And that's the best you can get? Real goals for and real goals against do not mean anything to you?

PDO is far, far from a perfect stat and guys like Laine are expected to ride "unsustainable" PDO as well as having hugely more real goals than those very imperfect xGoals model.

So in reality those lines worked really well, but in virtual reality you were right. The only problem is that only real goals count. Virtual ones doesn't. That's why winning the Corsi is not the goal of the game. Quality over quantity.

 

[Maukkis]

Oh noes. And that's the best you can get? Real goals for and real goals against do not mean anything to you?

PDO is far, far from a perfect stat and guys like Laine are expected to ride "unsustainable" PDO as well as having hugely more real goals than those very imperfect xGoals model.

So in reality those lines worked really well, but in virtual reality you were right. The only problem is that only real goals count. Virtual ones doesn't. That's why winning the Corsi is not the goal of the game. Quality over quantity.

So, you think that Hellebuyck turns into a .958 goalie behind those three, when they keep bleeding (high danger) shots against like no tomorrow?

That goal differential you keep clinging to for dear life is a mirage of unsustainable shooting and goaltending. The line performed very badly, but luck bailed them out and made the results look entirely different than what should have been the case. A larger sample size would, in all likelihood, prove you wrong, seeing as the stats that best predict future success indicate that the line performed poorly. That line is a great litmus test for seeing who among us understand probability and randomness, and you seem to struggle.

Here's an illustration. You're playing a slots machine which gives a 90% return on investment. After your first ten spins, you're winning, despite the odds being stacked against you. Should you continue playing? Obviously not. The same goes for icing a line which can be expected to perform negatively; don't put them together, period.

 

[jepjepjoo]

So, you think that Hellebuyck turns into a .958 goalie behind those three, when they keep bleeding (high danger) shots against like no tomorrow?

That goal differential you keep clinging to for dear life is a mirage of unsustainable shooting and goaltending. The line performed very badly, but luck bailed them out and made the results look entirely different than what should have been the case. A larger sample size would, in all likelihood, prove you wrong, seeing as the stats that best predict future success indicate that the line performed poorly. That line is a great litmus test for seeing who among us understand probability and randomness, and you seem to struggle.

Here's an illustration. You're playing a slots machine which gives a 90% return on investment. After your first ten spins, you're winning, despite the odds being stacked against you. Should you continue playing? Obviously not. The same goes for icing a line which can be expected to perform negatively; don't put them together, period.

I agree that the line wasn't that good, but that comparision doesn't really work... The slot will always revert to 90% return with enough spins while pdo is dependant on the skill of players on ice and doesn't eventually revert to 1.000.

 

[Ukkosenjumala]

PDO is a crude tool to predict whether or not something is under or over-performing but usually only in the more extreme ranges considering the smaller fluctuations can be accounted for by other things. Considering that CWL didn't play together more than a little over 100 minutes 5v5, it's entirely possible that they would have taken a nose-dive tho that can be attributed to many things, most importantly Wheeler not being a real center and both Laine and Connor being near-rookies with somewhat poor defensive game. Most of that anomaly in the PDO is that they were shooting at 15 %, something which I imagine would have Laine's footprint on it

This seems to go back to most of the lines that Laine has been succesful with, interesting to note that Ehlers-Stastny-Laine was also shooting at over 13 % as a line. They dominated goals for but again, Helleybuck was putting up almost 95 sv%. That whole 2017-2018 season is a bit skewed for PDO because of Helleybucks numbers, then again that's part of what makes PDO tricky because when you look at ELL, Helleybuck had a 90,6 sv% yet that line did about the same in corsi and high danger chances as ESL.

 

[kelsier]

Was planning to reply on earlier comments about a post of mine on another thread which incidentally seems to be closed as for the moment. Anyhow, I might as well continue here since I was addressing the line combo issues, particularly in the second line.

ffh: I never stated that the chemistry was anyone's fault. I mean how could it be? If two players don't mesh well together then they really should have no business whatsover playing together. As for Little "covering" Laine's mistakes, are you kidding me? I suppose with that statement you were referring to the defensive zone issues. Should we turn it around, would you mind to elaborate on who is covering Little's inability to create space and offence for his extremely talented wingers? With that said, would you rather have a) a line that does score goals while at the same time gets scored against b) a line that doesn't really score goals and gets scored against? Kind of mind-blowing I even have to present this question.

Rabid Ranger: Yeah, he most definitely knows what he is getting from Scheifele and Wheelers, but he should certainly know what he would be getting from Scheifele & Laine and Little - Wheelers. We already have this data from 16-17 and from beyond. The latter pair of Wheelers and Little have a long history together so Maurice should absolutely be aware how well they've worked with each other in the past. Here's the data jepjepjoo presented earlier:

Scheifele-Wheeler 2.93GF/60 2.41GA/60
Scheifele-Laine 4.20GF/60 2.80GA/60

Suppose it's needless to remind that as far as those numbers go, most of them are from the time when Laine was 18 years of age and he is now 20, and a much better all around player (far from a finished product but nonetheless). Let alone Scheifele having developed his game even further. I have to ask since you were pretty vocal about this matter, which one of these combos seem better to you exactly with the numbers presented and why so?

Now I wasn't even originally suggesting that Laine should be playing in the first line instead of Wheelers, but I did point out that the last games have been largely where getting points have been a matter of one or two goals and it's trending in a way that the Jets could start losing games while keeping inefficient second line intact and having a first line that doesn't really produce much either compared to other top units in the league. 5vs5 play certainly is a rather great concern going forward.

The fans absolutely have every right to question these decisions made by the management where all the avenues have not really been explored well, if at all. They have the depth which so far has managed to get points from most games but as far as line utilization goes and that being mirrored against the data presented with different line combinations from the past, the results could/should be even better. Some already mentioned earlier but also for instance we have Roslovic, who is a natural center and yet we have no real knowledge of him being ready to play in the middle and/or in the top six. Why? Never been tried out. Furthermore, what are they going to do this time around when the playoffs arrive with half-capable first line and utterly useless second line? Having a great powerplay goes only so far and the teams are much, much more careful with taking unnecessary penalties. Lets not forget that the Jets no longer have Stastny around to provide the experience and the capability to be that corner stone who't not only providing the much needed playmaking, but also handling the defensive duties required from a top six center for a contending team.

To think of it, Stastny's arrival and impact seemed kind of insane considering that the guy isn't even one of the better centers out there (good but not great), yet still, with talented players around him was able to provide pretty crazy impact and results. He did all that with much, much better efficiency than Little was and currently is able to provide. I find it odd how quickly people tend to forget this and instead circle back to where the whole discussion begun and continue like it never existed.

 

[LowLefty]

The first line is fine -

The second line?
KC and Patty need to get on their horse and start playing up to their ability -
We've seen this line perform earlier - why not now? What's different or has changed with this trio?

I'd work on finding an answer to that question before I hauled out the blender.
Besides, it's not going to happen anyway - and calling it out a couple a dozen times will not improve the chances of it happening.

Work on the issues with individual players and the line issues will work out themselves.
Just a theory . . .

 

[Jack722]

I'd like to see:

Perreault - Little - Laine
Give Laine another hustler to show him how it's done and give Lits some more checking/retrieval support. Perreault historically meshes well with Laine, which makes sense: Perreault creates chances and Laine buries them.

Connor - Roslovic - Petan/Appleton
I'd prefer Petan at centre but either configuration works. Surely this would be a line Maurice could trust with some minutes... right? Would love to see a super offense line like this to crush other teams bottom 6's.

Then reunite Copp-Lowry-Tanev.

 

[10Ducky10]

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

 

[Channelcat]

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

I like this. Petan? Idk, but I like Little with Wheeler. Little hasn't had chemistry with anyone since he was with Wheeler and Ladd, yet somehow Wheeler is the only top 6 winger he doesnt play with. And we all know 55 and 29 will be together eventually.

 

[voyageur]

All I know is that our lineup looked better before Christmas. Not sure bringing Copp in and taking Petan out has had the desired effect.

I would like to see the Moose line:
Petan-Roslovic-Appleton

Keep Perreault with Lowry-Tanev, adds the skill and playmaking element, without compromising tenacity.

I'd keep Schweelers together. I'd also keep Connor-Little-Laine together, but I'd tinker with Perreault on the wing. I'd also consider putting Laine on the 4th to fire him up, promote Roslovic or Appleton.

 

[Whileee]

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

 

[garret9]

Go ahead and explain then. I think you were suggesting they were a bad line due to PDO after others showed you they did alright in actual results. What did I get wrong?

Experiment...

You split half way through the season every player or line into 4 groups:

One with bad Corsi and good GF/GA + PDO.
One with bad Corsi and bad GF/GA + PDO.
One with good Corsi and good GF/GA + PDO.
One with good Corsi and bad GF/GA + PDO.

When both variables say the same thing, that’s boring... so ignore the those two groups.

Now with the two groups where both variables disagree... which one does better for the remainder of the season in out scoring?

It’s the good Corsi bad PDO group. They will have a better goal differential in the second half as a group than the group that had a better goal differential in the first half.

So when someone says something like “actual results” or “results that matter,” I like to counter with two things:

I care more about how someone will out score in the future vs how someone has out scored in the past.

The actual results that matter the most to me in evaluating players and lines are the ones that tell me how the team will play for the remainder of the year.

After all, putting someone on a line where they will no longer outscore just because they did doesn’t help much.

 

[Psych0dad]

Experiment...

You split half way through the season every player or line into 4 groups:

One with bad Corsi and good GF/GA + PDO.
One with bad Corsi and bad GF/GA + PDO.
One with good Corsi and good GF/GA + PDO.
One with good Corsi and bad GF/GA + PDO.

When both variables say the same thing, that’s boring... so ignore the those two groups.

Now with the two groups where both variables disagree... which one does better for the remainder of the season in out scoring?

It’s the good Corsi bad PDO group. They will have a better goal differential in the second half as a group than the group that had a better goal differential in the first half.

So when someone says something like “actual results” or “results that matter,” I like to counter with two things:

I care more about how someone will out score in the future vs how someone has out scored in the past.

The actual results that matter the most to me in evaluating players and lines are the ones that tell me how the team will play for the remainder of the year.

After all, putting someone on a line where they will no longer outscore just because they did doesn’t help much.

But past results are realized, not predictions based on averages. Which makes one of them accurate up to that point, and the other one can easily be inaccurate going forward because it is a prediction. There's at least a large margin of error in it.

I get that we can't expect past results to repeat themselves so it's not entirely accurate either, the performance will fluctuate. In young players cases mostly goes up and when aged past physical prime it goes down.

And then there's the individuals who break the average based models. They can't be accurately predicted by league averages. The more average the plauer is, the better the league averages can predict them.

Btw, would be interesting to hear your take on the line combos, what do the statistics tell you about how to optimize positive output?

 

[Maukkis]

But past results are realized, not predictions based on averages. Which makes one of them accurate up to that point, and the other one can easily be inaccurate going forward because it is a prediction. There's at least a large margin of error in it.

I get that we can't expect past results to repeat themselves so it's not entirely accurate either, the performance will fluctuate. In young players cases mostly goes up and when aged past physical prime it goes down.

And then there's the individuals who break the average based models. They can't be accurately predicted by league averages. The more average the plauer is, the better the league averages can predict them.

Btw, would be interesting to hear your take on the line combos, what do the statistics tell you about how to optimize positive output?

It would be ever so great if you showed that in your posts.

 

[garret9]

But past results are realized, not predictions based on averages. Which makes one of them accurate up to that point, and the other one can easily be inaccurate going forward because it is a prediction. There's at least a large margin of error in it.

I get that we can't expect past results to repeat themselves so it's not entirely accurate either, the performance will fluctuate. In young players cases mostly goes up and when aged past physical prime it goes down.

And then there's the individuals who break the average based models. They can't be accurately predicted by league averages. The more average the plauer is, the better the league averages can predict them.

Yes and no. Be careful as sometimes you are making erroneous leaps in conclusions...

"Accurate up to that point"

More accurate, in only particular context only...
Accurate that they did outscore: yes.
Accurate that they will outscore: no.
It told you what did happen, while other things tell you what will happen. The accuracy is only particular to a set of parameters. The answer to one question does not answer all questions.

Performance with age curves

Underlying statistics and scoring follows an aging curve; however, PDO does not. PDO predominately varies around 100 for all ages.

Individual sh% is different than on-ice sh%. You cannot conflate the two. Laine is a generational talent with the former, but in the latter the best guess is Laine's true talent is only worth 2 goals per 100 shots more than average.

Models "based on averages"

The reason why I use average in the experiment is because "luck" (ie: random variance) doesn't predict "luck." Some players will have bad luck in both halves of the sample, as will some have good luck in both halves of the sample.

IE: You split the distribution of PDO after the first half of the season at the mean and then look at each group separately for their second halves, both groups will have normal distributions around league mean PDO.

Goals have a huge amount of "luck" (random variance) in them. You are predicting a target that itself is a good chunk random. That's why you have to use averages. The average though can still be applied to the non-average (see later); that's two different things.

"Individuals that break the average based models"

No one "breaks" any statistical model to the point where PDO as a tool differs in how it should be used. All players regress to the mean, what is confusing you is that each player's mean is slightly different.

For all players a high or low PDO still means that their performance in goal is unsustainable. What differs is where their true talent PDO lies, although it's always much closer to the league mean than many people mistakenly tend to think.

IE: Even for Laine, you would look at him in different situations of high PDOs and low PDOs as unsustainable. His true talent may be closer to 102 than 100, but when you see a line where he posts 115 that's not skill/talent/chemistry... that's small sample variance.

Small samples variation

In an 82 game season, with thousands of shots, what makes one team different in sh% from the mean is about 2/3 "luck" (random variance), and 1/3 combination of shot quality (location, rebounds, etc.) and finishing talent combined.

Think about how much smaller samples you are going with when looking at Laine's different partners, where you are looking at extremely low number of shots and minutes with very different situations (goalies, linematching, score, etc.).

Laine and shot metrics

Shot metrics still matter for Laine. He scores on a greater number of shots than expected due to elite finishing skill, but that just shifts the model, not flips it upside down or makes it obsolete.

This makes logical sense. When Laine is on the ice you still want more chances and still want the opposition to have fewer. Having Laine or any other elite finisher only just shifts the translation factor.

Laine's best year in xG differential is also the same year as his best year in goal differential (2017-18). Laine's worst year in xG differential is also the same year as his worst year in goal differential (2018-19).

Laine doesn't break models. Laine shifts models.

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)

 

[garret9]

^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.

 

[EXTRAS]

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

 

[Keystone]

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

JoMo Trouba
Kulikov Myers
Chairot Morrow

My guess. Not horrendous if Myers can step up.