The Ralph Krueger put Jeff Skinner on the 4th line discussion including an excerpt from the instigators thread

[IKnowPhysics]

I was tooling around with WOWY to begin to understand Skinner's usage.  Skinner has played with centers Eichel, Staal, and Lazar.  The sample sizes are small, but here's the 5v5 result so far:

Center - TOI - CF%

Eichel - 4:51 - 88.89% (8CF 1CA)

Staal - 12:49 - 62.50% (15CF 9CA)

Lazar - 59:55 - 51.09% (47CF 45CA)

 

The obvious shines: Skinner is better with Eichel and Staal (and others) than Lazar.  Let's isolate Skinner and Lazar:

Both Skinner AND Lazar on ice - 61:34 - 51.06% (48CF 46CA)

Neither Skinner NOR Lazar on ice - 258:34 - 52.66% (228CF 205CA)

Skinner on ice without Lazar - 21:04 - 71.43% (25CF 10CA)

Lazar on ice without Skinner - 10:46 - 46.67% (7CF 8CA)

What you see is still obvious: Skinner generates more offense away from Lazar (because he's paired with Eichel or Staal and better RWs), and when they're together, Skinner lifts Lazar a little bit (but maybe within errors).  You may look at the second case and say, hey, the team is better without that pairing... does that mean they suck?  The answer is technically yes... but not for the reason you think.  The next part shows how the team is doing, in sorts.

 

Here's Eichel's WOWY so far this year:

Eichel on the ice - 112:38 - 60.21% (115CF 76CA)

Eichel not on the ice - 239:20 - 50.00% (193CF 193CA)

The team is even with Eichel off the ice and Eichel's line is murdering other teams 5v5.  Skinner and Lazar are 51.06%, which is slightly better than the non-Eichel average of 50.00%.  This sounds bleh, but this is an ENORMOUS improvement over the previous years where Eichel reigns and the rest of the lines get their ***** kicked in.  But last year, Eichel didn't exactly reign - he eked by:

Eichel on the ice 2019-20 - 1122:51 - 50.90%

Eichel not on the ice 2019-20 - 2270:16 - 48.20%

Improving the top line's CF% from 50.90% to 60.21% is absolutely massive, especially when they are recording a full third of the 5v5 minutes.  This must be a credit to Eichel or Hall or Krueger or all three, yes?  No.  Thank Sam Reinhart.  Here's why:

Hall and Eichel and no Reinhart - 25:01 - 51.35% (19CF 18CA) 

Hall and Eichel and Reinhart - 69:52 - 62.99% (80CF 47CA)

Once Sam is added to the top line, Corsi goes off the rails in our favor.  Our top line is outshooting opponents damn near two to one.  If you're Krueger, do you upset this machine?  Maybe it's worth it to try Skinner up there?  I mean, Skinner had those five minutes when Hall was out and they did pretty good...?  No.

 

Even though the non-Eichel Sabres are at CF% 50.00%, the High Danger Corsi (based on high danger scoring chances not just shot attempts) for the team shows that they're actually doing a great job at creating high danger scoring chances more than the opponent and that the shots they're generating are much high quality than the shots they're allowing:

Eichel on the ice - HDCF% 78.5%

Eichel not on the ice - HDCF% 58.5%

Skinner and Lazar are at HDCF% 62.50%, which is considerably better than the team without Skinner or Lazar or Eichel at HDCF% 53.19%.  Skinner is helping the Lazar line tilt the ice in the dangerous areas and he's doing a better job at it than anyone not playing with Eichel.

[dudacek]

16 minutes ago, IKnowPhysics said:

I was tooling around with WOWY to begin to understand Skinner's usage.  Skinner has played with centers Eichel, Staal, and Lazar.  The sample sizes are small, but here's the 5v5 result so far:

Center - TOI - CF%

Eichel - 4:51 - 88.89% (8CF 1CA)

Staal - 12:49 - 62.50% (15CF 9CA)

Lazar - 59:55 - 51.09% (47CF 45CA)

 

The obvious shines: Skinner is better with Eichel and Staal (and others) than Lazar.  Let's isolate Skinner and Lazar:

Both Skinner AND Lazar on ice - 61:34 - 51.06% (48CF 46CA)

Neither Skinner NOR Lazar on ice - 258:34 - 52.66% (228CF 205CA)

Skinner on ice without Lazar - 21:04 - 71.43% (25CF 10CA)

Lazar on ice without Skinner - 10:46 - 46.67% (7CF 8CA)

What you see is still obvious: Skinner generates more offense away from Lazar (because he's paired with Eichel or Staal and better RWs), and when they're together, Skinner lifts Lazar a little bit (but maybe within errors).  You may look at the second case and say, hey, the team is better without that pairing... does that mean they suck?  The answer is technically yes... but not for the reason you think.  The next part shows how the team is doing, in sorts.

 

Here's Eichel's WOWY so far this year:

Eichel on the ice - 112:38 - 60.21% (115CF 76CA)

Eichel not on the ice - 239:20 - 50.00% (193CF 193CA)

The team is even with Eichel off the ice and Eichel's line is murdering other teams 5v5.  Skinner and Lazar are 51.06%, which is slightly better than the non-Eichel average of 50.00%.  This sounds bleh, but this is an ENORMOUS improvement over the previous years where Eichel reigns and the rest of the lines get their ***** kicked in.  But last year, Eichel didn't exactly reign - he eked by:

Eichel on the ice 2019-20 - 1122:51 - 50.90%

Eichel not on the ice 2019-20 - 2270:16 - 48.20%

Improving the top line's CF% from 50.90% to 60.21% is absolutely massive, especially when they are recording a full third of the 5v5 minutes.  This must be a credit to Eichel or Hall or Krueger or all three, yes?  No.  Thank Sam Reinhart.  Here's why:

Hall and Eichel and no Reinhart - 25:01 - 51.35% (19CF 18CA) 

Hall and Eichel and Reinhart - 69:52 - 62.99% (80CF 47CA)

Once Sam is added to the top line, Corsi goes off the rails in our favor.  Our top line is outshooting opponents damn near two to one.  If you're Krueger, do you upset this machine?  Maybe it's worth it to try Skinner up there?  I mean, Skinner had those five minutes when Hall was out and they did pretty good...?  No.

 

Even though the non-Eichel Sabres are at CF% 50.00%, the High Danger Corsi (based on high danger scoring chances not just shot attempts) for the team shows that they're actually doing a great job at creating high danger scoring chances more than the opponent and that the shots they're generating are much high quality than the shots they're allowing:

Eichel on the ice - HDCF% 78.5%

Eichel not on the ice - HDCF% 58.5%

Skinner and Lazar are at HDCF% 62.50%, which is considerably better than the team without Skinner or Lazar or Eichel at HDCF% 53.19%.  Skinner is helping the Lazar line tilt the ice in the dangerous areas and he's doing a better job at it than anyone not playing with Eichel.

Wait, you’re saying that Ralph’s lineup choices are actually supported by analytics?

 

[inkman]

Amazing post @IKnowPhysics

[Zamboni]

The Reinhart “haters” hate those facts. Ignore the facts. And come up with eye test “data” or random and completely false things like “slow skater” or “doesn’t do any one thing really really like super duper great”, or my favorite one “doesn’t go in the dirty areas”. 😂

[steveoath]

How do Skinners fancy stats compare with those of Olofsson. As I see it that's the two that are "competing" for 1 top 6 place. Eye test seems to suggest Victor has improved team play, which would play into the coaches plans of synergy (and other management craptalk).,

[PerreaultForever]

3 hours ago, Zamboni said:

The Reinhart “haters” hate those facts. Ignore the facts. And come up with eye test “data” or random and completely false things like “slow skater” or “doesn’t do any one thing really really like super duper great”, or my favorite one “doesn’t go in the dirty areas”. 😂

Maybe but I'm too old to start watching hockey like a mathematician. Where's the fun in that? We old timers might not bother to try to understand all these fancy dancy numbers but we've seen winning hockey and know what it takes to win. And for you youngsters believe it or not, we've seen winning Sabres hockey too lol. 

many a night those stats have said Sam was good but that top line was held scoreless. Go figure. 

and no, he DOESN'T go into the dirty areas. 🙂

[Taro T]

9 hours ago, IKnowPhysics said:

I was tooling around with WOWY to begin to understand Skinner's usage.  Skinner has played with centers Eichel, Staal, and Lazar.  The sample sizes are small, but here's the 5v5 result so far:

Center - TOI - CF%

Eichel - 4:51 - 88.89% (8CF 1CA)

Staal - 12:49 - 62.50% (15CF 9CA)

Lazar - 59:55 - 51.09% (47CF 45CA)

 

The obvious shines: Skinner is better with Eichel and Staal (and others) than Lazar.  Let's isolate Skinner and Lazar:

Both Skinner AND Lazar on ice - 61:34 - 51.06% (48CF 46CA)

Neither Skinner NOR Lazar on ice - 258:34 - 52.66% (228CF 205CA)

Skinner on ice without Lazar - 21:04 - 71.43% (25CF 10CA)

Lazar on ice without Skinner - 10:46 - 46.67% (7CF 8CA)

What you see is still obvious: Skinner generates more offense away from Lazar (because he's paired with Eichel or Staal and better RWs), and when they're together, Skinner lifts Lazar a little bit (but maybe within errors).  You may look at the second case and say, hey, the team is better without that pairing... does that mean they suck?  The answer is technically yes... but not for the reason you think.  The next part shows how the team is doing, in sorts.

 

Here's Eichel's WOWY so far this year:

Eichel on the ice - 112:38 - 60.21% (115CF 76CA)

Eichel not on the ice - 239:20 - 50.00% (193CF 193CA)

The team is even with Eichel off the ice and Eichel's line is murdering other teams 5v5.  Skinner and Lazar are 51.06%, which is slightly better than the non-Eichel average of 50.00%.  This sounds bleh, but this is an ENORMOUS improvement over the previous years where Eichel reigns and the rest of the lines get their ***** kicked in.  But last year, Eichel didn't exactly reign - he eked by:

Eichel on the ice 2019-20 - 1122:51 - 50.90%

Eichel not on the ice 2019-20 - 2270:16 - 48.20%

Improving the top line's CF% from 50.90% to 60.21% is absolutely massive, especially when they are recording a full third of the 5v5 minutes.  This must be a credit to Eichel or Hall or Krueger or all three, yes?  No.  Thank Sam Reinhart.  Here's why:

Hall and Eichel and no Reinhart - 25:01 - 51.35% (19CF 18CA) 

Hall and Eichel and Reinhart - 69:52 - 62.99% (80CF 47CA)

Once Sam is added to the top line, Corsi goes off the rails in our favor.  Our top line is outshooting opponents damn near two to one.  If you're Krueger, do you upset this machine?  Maybe it's worth it to try Skinner up there?  I mean, Skinner had those five minutes when Hall was out and they did pretty good...?  No.

 

Even though the non-Eichel Sabres are at CF% 50.00%, the High Danger Corsi (based on high danger scoring chances not just shot attempts) for the team shows that they're actually doing a great job at creating high danger scoring chances more than the opponent and that the shots they're generating are much high quality than the shots they're allowing:

Eichel on the ice - HDCF% 78.5%

Eichel not on the ice - HDCF% 58.5%

Skinner and Lazar are at HDCF% 62.50%, which is considerably better than the team without Skinner or Lazar or Eichel at HDCF% 53.19%.  Skinner is helping the Lazar line tilt the ice in the dangerous areas and he's doing a better job at it than anyone not playing with Eichel.

Thanks for that, IKP.  As someone else mentioned, would be very interesting to see Olofsson's #'s.  (If you have a free hour or so you want to kill. 😉 )  

And your data supports the eye test.  Eichel & Hall were getting chances w/ Olofsson.  They've been getting ridiculous amounts of chances w/ Reinhart.  Have to believe it's just a matter of time for them to start potting them.  (Skinner w/ his chances, too.)

One item not in your data, but from the eye test: Eichel's line had difficulty getting the puck out of their end w/ Dahlin-Montour & now have even more getting the puck up ice w/ Dahlin-Miller.  (But once they were out of their zone, lookout.)

[Zamboni]

4 hours ago, PerreaultForever said:

Maybe but I'm too old to start watching hockey like a mathematician. Where's the fun in that? We old timers might not bother to try to understand all these fancy dancy numbers but we've seen winning hockey and know what it takes to win. And for you youngsters believe it or not, we've seen winning Sabres hockey too lol. 

many a night those stats have said Sam was good but that top line was held scoreless. Go figure. 

and no, he DOESN'T go into the dirty areas. 🙂

Age has nothing to do with it… And I’m not a youngster… At all. and that’s where you and I differ and will always differ … Regarding Reinhart. I see him go to the net, stand around the net, go in the corners to dig out the puck. I see him take a beating some nights. You do not see it obviously. So thanks for showing your bias regarding Sam once again. You and I both know it’ll never change. you should really stop trying to hide behind age without knowing who you’re talking to. There’s a difference between knowing stats, knowing how to look at stats, understanding stats… And then just flat out refusing to believe the data.  

[JohnC]

1 hour ago, Zamboni said:

Age has nothing to do with it… And I’m not a youngster… At all. and that’s where you and I differ and will always differ … Regarding Reinhart. I see him go to the net, stand around the net, go in the corners to dig out the puck. I see him take a beating some nights. You do not see it obviously. So thanks for showing your bias regarding Sam once again. You and I both know it’ll never change. you should really stop trying to hide behind age without knowing who you’re talking to. There’s a difference between knowing stats, knowing how to look at stats, understanding stats… And then just flat out refusing to believe the data.  

There isn't a player who is more willing to go to the net (dirty area) and absorb a pummeling than Reinhart. One of the noticeable deficiencies last year was the lack of net presence. It's a little better this year but there is still a  need for more of it.  As a general observation Reinhart is one of our tougher and grittier players. It's not as noticeable as some of the bigger more open ice hitters but it is very much there. 

 

[LGR4GM]

So basically what I am seeing is that Skinner is being completely misused by Krueger. 

[dudacek]

1 minute ago, LGR4GM said:

So basically what I am seeing is that Skinner is being completely misused by Krueger. 

That's what you got from IKP's post?

[JohnC]

1 minute ago, LGR4GM said:

So basically what I am seeing is that Skinner is being completely misused by Krueger. 

Not completely misused but handled with a little more flexibility. When you have an asset that doesn't quite fit your preferred mold a little more forbearance might be appropriate.

[LGR4GM]

11 minutes ago, dudacek said:

That's what you got from IKP's post?

Yes. Also I looked at Olofsson's numbers, he's better but Skinner is better still. 

[dudacek]

3 minutes ago, LGR4GM said:

Yes. Also I looked at Olofsson's numbers, he's better but Skinner is better still. 

Please explain what his numbers show we should be doing.

I'm seeing a strategy that is allowing all four lines to outchance their opposition.

[LGR4GM]

41 minutes ago, dudacek said:

Please explain what his numbers show we should be doing.

I'm seeing a strategy that is allowing all four lines to outchance their opposition.

Is that what we are trying to do? Out-chance the other team? 

[Taro T]

15 minutes ago, LGR4GM said:

Is that what we are trying to do? Out-chance the other team? 

Obviously, the goal is to outscore the other team.  When your goaltending is only middling, if you aren't outchancing the opposition, you are likely losing more than you are winning.

[LTS]

26 minutes ago, LGR4GM said:

Is that what we are trying to do? Out-chance the other team? 

To a certain degree that would be the case would it not?

If the goal is to maximize the output of all 4 lines then improving one line at the expense of another might not make sense.  it really depends.  You move Skinner, does that decrease the line that he moved away from?

And as Taro said, naturally your goal is to outscore the other team.  Theoretically that comes from maximizing your opportunities to score.

Ultimately, all of these stats mean nothing if they don't win.  Anyone can talk about "regression to the mean" but if that never happens someone has to go find the reason why it didn't.. because it means their model was flawed.  I understand that the stats are used to try and create the use of resources that support the goal of winning and so they are important.

The total and final measuring stick is simply this.  Did you win enough to win it all? 

[IKnowPhysics]

Thanks for the positive feedback.  It's nice to know I'm not yelling equations into a void.

I get where LTS is going, but there's just not enough data to know if the alternative is better.  But there's no data if Krueger doesn't try it, so it's a chicken and egg problem.

But I get it.

To win, you need to score goals and prevent goals.

To score goals, you need lots of shots multilplied by a good shooting percentage.  To prevent goals, you need to prevent shots multiplied by forcing shots to have a low shooting percentage.

You'd like to simultaneously win CF% with margin and have great SH% and great SV%, but realistically, you're trying to win the overall HDCF% battle, as high danger corsi typically has a much higher SH% and lower SV% against and these chances make bigger impacts on wins/losses over the season than low percentage plays.

LTS points out that it's unknown whether the overall team HDCF when Skinner is with Lazar is any better or worse than when Skinner is with Eichel or Staal (or others).  Maybe Skinner-Eichel-Reinhart is better than Hall-E-R and Hall aids Lazar just the same as Skinner, causing an overall team improvement in HDCF%.  Hard to know unless it's tried.

 

[LabattBlue]

HC's always have a double standard when it comes to "performance determines play".  Early into this season, but Dahlin has arguably been the worst Sabre on some nights, yet he still dresses every game, still plays significant minutes, and still gets PP time.  So for all those who say Skinner deserves what he gets for not buying in, I say BS.

Edited January 28, 2021 by LabattBlue

[erickompositör72]

8 minutes ago, LabattBlue said:

HC's always have a double standard when it comes to "performance determines play".  Early into this season, but Dahlin has arguably been the worst Sabre on some nights, yet he still dresses every game, still plays significant minutes, and still gets PP time.  So for all those who say Skinner deserves what he gets for not buying in, I say BS.

And on the next episode of apples to oranges...

 

😜

[PerreaultForever]

6 hours ago, Zamboni said:

Age has nothing to do with it… And I’m not a youngster… At all. and that’s where you and I differ and will always differ … Regarding Reinhart. I see him go to the net, stand around the net, go in the corners to dig out the puck. I see him take a beating some nights. You do not see it obviously. So thanks for showing your bias regarding Sam once again. You and I both know it’ll never change. you should really stop trying to hide behind age without knowing who you’re talking to. There’s a difference between knowing stats, knowing how to look at stats, understanding stats… And then just flat out refusing to believe the data.  

Dude, I'm not "hiding."  I watch hockey with friends and none of us talk about corsi and fenwick and all that stuff cause it's just not interesting or fun to look at team sports that way. Go play videogames if you want to look at everything in a dissected analytical way. We watch sports for fun and entertainment. Analytic worship tends to be an age thing, but if you are old and into stats too good for you. Pretty sure you're in a minority view for your age cohort but it DOESN'T MATTER. Enjoy your thing and I'll enjoy mine Ok?

Now as for Sam, the eyes don't lie. The stat will tell you he was in that part of the ice but it won't tell you that he backed away at the moment the other guy gave him a stern look. (Obviously exaggerating there to make a point, but I think you know what I mean). By the way, it's not a bias. He's played better this season than usual imo. Better than Eichel at times. That's what my eyes say.

I'm not going to do this, cause I'm not into stats, but since you are I have a challenge for you out of curiosity. Take Sam's stats and run a comparison with other wingers who clearly do go to the dirty areas often. Tell me how they stack up. I'd suggest Marchand. Maybe Tkachuk. Somebody like that. How do they stack up? Cause if Sam's stats say he's in there as much as those guys, then the stats clearly don't tell you everything they need to.

[dudacek]

I'd say Sam does into the dirty areas as much as Marchand and Tkachuk.

He just doesn't go in with his stick up and his gums flapping.

[PerreaultForever]

1 hour ago, LabattBlue said:

HC's always have a double standard when it comes to "performance determines play".  Early into this season, but Dahlin has arguably been the worst Sabre on some nights, yet he still dresses every game, still plays significant minutes, and still gets PP time.  So for all those who say Skinner deserves what he gets for not buying in, I say BS.

I agree with this. Dahlin needs to sit a few games. 

When Mcavoy was struggling in Boston I think it was 2 seasons back they benched and then sat him out and it caused rumblings there about him not being as good as first thought. Woke him up or he took a step back or whatever and they paired him with a veteran (to say the least) and now, in the bruins game I watched this year he looked like a Norris candidate. 

We have a history of developing D men poorly and we might be doing it again. I still think this falls on Steve Smith and he should have gone with Housley. But whatever, it's time he sat. 

[LGR4GM]

3 hours ago, LTS said:

To a certain degree that would be the case would it not?

If the goal is to maximize the output of all 4 lines then improving one line at the expense of another might not make sense.  it really depends.  You move Skinner, does that decrease the line that he moved away from?

And as Taro said, naturally your goal is to outscore the other team.  Theoretically that comes from maximizing your opportunities to score.

Ultimately, all of these stats mean nothing if they don't win.  Anyone can talk about "regression to the mean" but if that never happens someone has to go find the reason why it didn't.. because it means their model was flawed.  I understand that the stats are used to try and create the use of resources that support the goal of winning and so they are important.

The total and final measuring stick is simply this.  Did you win enough to win it all? 

what if leaving skinner where he is gives that line a +4% in overall corsi, say instead of 47 they are 51. But let's say I move him to a line that has a 56% corsi now and they go up to a 62% because Cozens and Staal work well with Skinner... I gained 2% overall but I am supposed to believe this was bad?

[Zamboni]

PerFor ... I’m absolutely not going to waste my time with your “wish list” of forwards I should look up stats on. You and I both know it’s a waste of time. You have your bias mind made up regarding Reinhart. Nothing I or others say today will change your mind. Ever. It’s a blind spot you have. Over the years, posters have done a “deep dive” into Reinhart with and without certain line mates. And almost EVERY time it shows favorably in Reinharts favor. But no ... the eye test and “feelings” are what you hold as truth instead of the facts sometimes.

I also don’t need to explain to you, how I watch games. But it certainly isn’t like a video game. And it certainly isn’t being immersed in stats or analytics as the game is happening, besides the analytics that are presented during the intermissions that any 12-year-old who follows hockey can understand. I’ll look at the general stats and basic analytics after games. And it’s usually only the stats and analytics that are presented from posters on this forum. I look at other sites as well don’t get me wrong… But not every single game. If you choose to ignore any information except for what your eyes and feelings tell you… Then that’s the way you choose to enjoy the game. Good for you. You refuse to except that sometimes your eyes and feelings can deceive you… Especially when analytics are slapping you in the face telling you that you’re wrong.

bottom line is… You and I both know it’s a waste of time for anybody to try to convince you that Reinhart is different than what you perceive. If somebody else would like to waste their time trying to once again convince you to think differently about Reinhart, they can. It won’t be me. but the “challenge“ that you proposed is an interesting one. But I think it should be extended to several other players and not Tkachuk or the scumbag Marchand.