Player Discussion: Winnipeg Jets Defense

[10Ducky10]

The one and only reason Boolow is playing is because of the PK.
I hope Slamberg is seeing LOTS of time on the PK with the Moose.

 

[voyageur]

All your looking for is someone to completely agree with you.

Beaulieu is garbage. Can he be a #7 who steps in for injuries? Sure, but he does not deserve a regular shift in the NHL and having him block guys like Stanley/Samberg/Heinola from playing is malpractice by coaching.

Interestingly when looking at statistics, our garbage defenseman has exactly the same ratio of GF/GA at 2.9/60 as Morrissey and Pionk. Now part of that came riding on good shifts from Copp-Stastny-Ehlers, but it's hard to discern where the garbage begins, and where the good players have separated themselves.

For the record, it's not the coach who signed 8 defensemen to start the year to NHL contracts.

Look at the current roster and when Thompson comes off IR there is not a single ELC who has room on the current roster to make the team.

Now the question you should ask is did Maurice make a mistake by playing Beaulieu over Stanley? And if so do you think he made a mistake playing Vesalainen, because the only way for Stanley to keep playing is for Vesalainen to be demoted to the taxi squad? And given that Stanley had his worst game of the season, on the road where he can't be matchup protected, is that a poor coaching decision.? Or do you think that Maurice should be actively canvassing our GM to have veterans or any players waived? Doesn't sound like the coach's job.

So then you see that the coach isn't the issue but the target of ignorance.

Maybe Chevy had a plan to develop his players, after drafting them. Seems like that has been the plan all along and Pascal Vincent has done an exemplary job of getting young players to the next level.

I loved this quote from Vincent:

“That’s the reason of our existence. Having young guys and to be able to play them. That’s what we do. At the end of the game and we have those young guys on the ice and we have Perfetti on a six-on-five, we put them in all situations, power play, penalty kill, that’s why we exist. We want to give them as much mileage as possible in order to help them grow. A pro-hockey game is different, it takes a 60 minute effort. It goes faster, you’re playing against men, and the more you do it you get used to it. To us, it’s really exciting to have so many young players. On defence, we talk about the age but the skills are there, they can play. They can make plays. They had real good breakouts, the gap was right very often, we need to improve there’s no doubt but I think they did well in their first game.”

For a team that is winning hockey games, and part of it is special teams, where a guy like Beaulieu is definitely contributing the right way, sure are a lot of sour grapes.

 

[Ducky10]

Not to far fetched,

Yes it is, and I like Poolman.

 

[JetsFan815]

We should have signed Troy Stecher in the offseason, I was hoping we'd trade for him since the last deadline. Off to a great start on Detroit. Would look good on our blueline playing in the BooBoo spot.

 

[voyageur]

We should have signed Troy Stecher in the offseason, I was hoping we'd trade for him since the last deadline. Off to a great start on Detroit. Would look good on our blueline playing in the BooBoo spot.

Except he is a RD. And we have our RD, and I think that each of Pionk, Poolman, and De Melo fill some kind of need. The first two bring a lot of mobility, De Melo is expected to be a defensive leader. Not sure Stetcher brings enough to knock any of those guys out of the lineup.

I think Chevy has got it right. I fully expect Niku to get waived or traded in the next week or so. That should open up a spot for Stanley to run with. And should he falter, Samberg could simply take it, while Stanley moves to the taxi squad.

While Morrissey-Poolman, Forbort-Pionk, Stanley-De Melo is not a sexy defense I think it fills out rather nice, and the biggest thing is to build chemistry over the course of the year.. Boo Boo as #7 gives us some depth.

If we are knocking Beaulieu out of the lineup, I think it can be done by giving Morrissey-Poolman the 2nd PK reps, as both had good metrics last year, and maybe Stanley gets eased in. This is an area where Samberg may contribute later in the season. However to make that possible I would suggest moving Pionk onto the 1st PP unit, and Morrissey to the 2nd, to balance out the minutes for the 2, as these should be the only d-men who regularly hit well over 20 minutes. Forbort and De Melo might hit the 20 minute mark in a game where we are in the box a lot, but frankly the Jets have been pretty good at not taking penalties.

I also think if Samberg is as good as predicted, he could potentially knock Forbort out of the top 4 at some point. Where maybe we use Morrissey-De Melo as a shutdown pair, run Samberg-Pionk, which would be a pretty good skating pairing, and Forbort-Poolman as a 3rd pairing.

I am not sure we needed to add anyone external, as much as wait for someone to grow internally.

 

[truck]

For a team that is winning hockey games, and part of it is special teams, where a guy like Beaulieu is definitely contributing the right way, sure are a lot of sour grapes.

Beaulieu is a below average penalty killer on a below average penalty kill.

I would have no problem with Mo holding spots for PK specialists if the PK wasn't consistently in the bottom half of the league.

His obsession with role players that fit a pre-determined "typed" IS the problem.

 

[garret9]

Utter and complete garbage?

He's moved from being on pace to be the 3rd to now being the 2nd worst defender according to xGoal +/- (RAPM adjusted) in all of NHL 2.0 history.

That's pretty bad.

FYI to others here: That's worse than any of Niku's 3 bad seasons.

 

[Joe Hallenback]

I wouldn't be surprised by the end of the season to see both Heinola and Samberg in the lineup. From what I have seen on the Moose they are both ready. Heck Kovacevic is ready to play as well and Chisholm looks real good too

 

[garret9]

Actually you corrected me well. I didn't realize that BooBoo spent so much time in the minors with the Habs, so my old memory served me wrong. I just remember he was a promising prospect right up until they acquired Weber, and he couldn't seize the top 4 job he was expected to. With Buffalo he played behind Mc Cabe and Scandella who are better players, Bogosian was still a top 4 RD man in the rotation. though Gorges was in the mix on both sides. When Dahlin was drafted, Beaulieu fell off the depth charts. Only when the Sabres acquired Montour and Jokiharju did Bogo fall off. They actually have pretty similar career path.

As for +/- last year, you can call it luck, but no one else was that lucky. And there's always the danger of measuring a player on shots against vs. goals against, and take 31 coaches I'm pretty sure which measurement they'll use. As for the PK, he's not bad at it. The team was terrible in general last year, and he was in that mediocre range. This year he has been better, in that aspect. Saying it's possible he could be worse, is speculative, because he's been steadily improving in that department. Regardless it's a moot discussion. I'm all for someone getting ahead of Beaulieu in the lineup if they deserve it. I'd say Stanley is on that path, and we'd be better off in turn with Beaulieu in the 7 spot, who's had some success with players on this roster (Stanley for one) as opposed to Niku, who's potential has not equated to results. Which is the worst place to be as a young player. Once we clear Niku, we could conceivably make room for Samberg, and then it's up to him to earn his stripes, and if he is good enough he'll knock both Stanley and Boo Boo out of the lineup permanently.

Actually, no one professionally in an executive position or coaching actually uses plus-minus other than for sound bites.

Plus-minus is now just used exclusively by some fans and media, and some players.

 

[garret9]

Yup. Teams are also soon going to be scanning for eligible D for the expansion draft.

I just think of it this way, Jets waiving Beaulieu or putting him in the pressbox and never letting him on the ice helps their on ice product by similar amount.

Keeping him on the roster though does help:
* expansion draft
* guy that some teammates like
* veteran presence bs

 

[voyageur]

Actually, no one professionally in an executive position or coaching actually uses plus-minus other than for sound bites.

Plus-minus is now just used exclusively by some fans and media, and some players.

Well I have never heard anyone professional besides yourself use xGoals +/- (RAPM adjusted) so I am going to take it with a grain of salt, because I think that some of the stats that coaches would rely on are real time and not cumulative. And not all of them are shot based either, especially for defense. As we saw with Maurice talking about kill plays. How do you measure one on one battles for instance, G, in your metrics?

Anyways it's a dead horse. I'm sure the same people that voted to fire Maurice would vote to fire Beaulieu off into the sun.

I'm optimistic about the future of our defense like anyone else. There's definitely room for improvement in the mobility department.

I'd like your take on why we give up so many scoring chances in the slot.

 

[garret9]

Well I have never heard anyone professional besides yourself use xGoals +/- (RAPM adjusted) so I am going to take it with a grain of salt, because I think that some of the stats that coaches would rely on are real time and not cumulative. And not all of them are shot based either, especially for defense. As we saw with Maurice talking about kill plays. How do you measure one on one battles for instance, G, in your metrics?

Anyways it's a dead horse. I'm sure the same people that voted to fire Maurice would vote to fire Beaulieu off into the sun.

I'm optimistic about the future of our defense like anyone else. There's definitely room for improvement in the mobility department.

I'd like your take on why we give up so many scoring chances in the slot.

You really want me to bring in receipts...

Okay here we go, guys that I know use RAPM or similar models and would suggest not using +/- for evidence of anything:
(CAR) - Eric Tulsky, Kevin Kan
(COL) - Arik Parnass, Dawson Sprigins (oh he brought RAPM to hockey)
(CBJ) - Josh Flyn
(CHI) - Stan Bowman
(FLA) - Cam Lawrence, Josh Weissbock
(LAK) - Rob Volman
(MIN) - Mat Sells
(NJD) - Matt Cane
(PIT) - Sam Ventura
(SEA) - Alexandra Mandrycky, Namita Nandakumar, Dani Chu
(TBL) - Michael Peterson
(TOR) - Kyle Dubas, Eric Joyce, Cam Charron, Daryl Metcalf
(VAN) - Jonathan Wall, Aidan Fox
(WSH) - Tim Barnes

That's just ones that I personally know that would say that RAPM and similar models have their utility and traditional +/- does not... there's more out there than that.
Some would point out that these models are imperfect (as would I, I just published a piece on Hockey-Graphs today about that), but at least the model answers questions. Traditional +/- just answers what was their +/-.

 

[DannyGallivan]

Interestingly when looking at statistics, our garbage defenseman has exactly the same ratio of GF/GA at 2.9/60 as Morrissey and Pionk. Now part of that came riding on good shifts from Copp-Stastny-Ehlers, but it's hard to discern where the garbage begins, and where the good players have separated themselves.

So much as to do with who the competition is when they're on the ice. Poor Pionk had the McDavid assignment last game. Morrissey always logs the most difficult minutes (at least he used to before Pionk and Forborg started taking off).

 

[DRW204]

You really want me to bring in receipts...

Okay here we go, guys that I know use RAPM or similar models and would suggest not using +/- for evidence of anything:
(CAR) - Eric Tulsky, Kevin Kan
(COL) - Arik Parnass, Dawson Sprigins (oh he brought RAPM to hockey)
(CBJ) - Josh Flyn
(CHI) - Stan Bowman
(FLA) - Cam Lawrence, Josh Weissbock
(LAK) - Rob Volman
(MIN) - Mat Sells
(NJD) - Matt Cane
(PIT) - Sam Ventura
(SEA) - Alexandra Mandrycky, Namita Nandakumar, Dani Chu
(TBL) - Michael Peterson
(TOR) - Kyle Dubas, Cam Charron, Daryl Metcalf
(VAN) - Jonathan Wall, Aidan Fox
(WSH) - Tim Barnes

That's just ones that I personally know that would say that RAPM and similar models have their utility and traditional +/- does not... there's more out there than that.
Some would point out that these models are imperfect (as would I, I just published a piece on Hockey-Graphs today about that), but at least the model answers questions. Traditional +/- just answers what was their +/-.

wonder what maurice and his supposed in-house analytics use

 

[DannyGallivan]

I wouldn't be surprised by the end of the season to see both Heinola and Samberg in the lineup. From what I have seen on the Moose they are both ready. Heck Kovacevic is ready to play as well and Chisholm looks real good too

That may be four fifths of our defensive line-up in three years. Don't forget Stanley, who is only getting better. Pionk and Morrissey may be the only leftovers from last game's starting D in three years.

... or maybe two years.

 

[garret9]

wonder what maurice and his supposed in-house analytics use

I know about many non-WPG teams more than I know about WPG.

I know they do import data from SportsLogiq, which also comes with models. I'm not a huge fan of SL's models although their data can be used to build good models. I don't know whether they use SL data to build their own or just use SL models though.

For staff, I know of 3 main members: Max Erenberg, Matt Prefontaine, and Larry Simmons

Simmons is the AGM. I've had breif discussions as a potential vendor when HockeyData was potentially selling data to Jets. I wasn't part of all those convos (I mostly just deal with IP for the company) but I did had brief talks to Cheveldayoff, Simmons, and Maurice during that time. I'm guessing of the 3 of them he uses the data the most, but I doubt any of the grunt work. My guess is his main usage of analytics is Capologist type stuff but I'm throwing a dart in the dark without any real knowledge.

Prefontaine is the Video Coach/Analyst for the Jets and deals with analytics as well. I don't know what his specific role is, if maybe it's more traditional where he just links video to analytics or more hybrid modern role. I've never had any hockey convos with him. I did talk to him once about bringing up a young family and he seemed super nice though.

Erenberg is the Jets only designated hockey analytics person that I know of. I have never met or talked with him. Our only contact is that we both follow each other on Twitter. That's it. I do know of people (some of the ones I listed above) who have talked to him, and they all said he was nice. But again, I don't know his standing on any theoretical or philisophical standings on modeling and other stuff to make a call.

 

[truck]

I wouldn't be surprised by the end of the season to see both Heinola and Samberg in the lineup. From what I have seen on the Moose they are both ready. Heck Kovacevic is ready to play as well and Chisholm looks real good too

I would be surprised to see them in the lineup. Save for a significant injury to someone like Morrissey or Pionk, I don't expect to see Heinola in the line-up.

He doesn't kill penalties, so Mo won't want him in a depth role.

 

[Weezeric]

I know about many non-WPG teams more than I know about WPG.

I know they do import data from SportsLogiq, which also comes with models. I'm not a huge fan of SL's models although their data can be used to build good models. I don't know whether they use SL data to build their own or just use SL models though.

For staff, I know of 3 main members: Max Erenberg, Matt Prefontaine, and Larry Simmons

Simmons is the AGM. I've had breif discussions as a potential vendor when HockeyData was potentially selling data to Jets. I wasn't part of all those convos (I mostly just deal with IP for the company) but I did had brief talks to Cheveldayoff, Simmons, and Maurice during that time. I'm guessing of the 3 of them he uses the data the most, but I doubt any of the grunt work. My guess is his main usage of analytics is Capologist type stuff but I'm throwing a dart in the dark without any real knowledge.

Prefontaine is the Video Coach/Analyst for the Jets and deals with analytics as well. I don't know what his specific role is, if maybe it's more traditional where he just links video to analytics or more hybrid modern role. I've never had any hockey convos with him. I did talk to him once about bringing up a young family and he seemed super nice though.

Erenberg is the Jets only designated hockey analytics person that I know of. I have never met or talked with him. Our only contact is that we both follow each other on Twitter. That's it. I do know of people (some of the ones I listed above) who have talked to him, and they all said he was nice. But again, I don't know his standing on any theoretical or philisophical standings on modeling and other stuff to make a call.

Quick question for you Garret, if you have a moment.

Do you have any thoughts on tracking? There seems to be some recent discussion about tracking errors from the NHL, aside from the subjective nature of some events. Is the tracking from the NHL reliable? Is there any concern that the data might be very different from one individual to the next? Does your company do your own?

 

[garret9]

Quick question for you Garret, if you have a moment.

Do you have any thoughts on tracking? There seems to be some recent discussion about tracking errors from the NHL, aside from the subjective nature of some events. Is the tracking from the NHL reliable? Is there any concern that the data might be very different from one individual to the next? Does your company do your own?

NHL's data is imperfect. Private companies is also imperfect but generally do much better.

That doesn't make the data for the NHL bad though.

Like private data might end up calling what public data says is a weaker 2nd line player to a stronger 2nd line player, but private data isn't flipping public data upside down saying the 4th liner is actually an all star or vice versa.

 

[DRW204]

I know about many non-WPG teams more than I know about WPG.

I know they do import data from SportsLogiq, which also comes with models. I'm not a huge fan of SL's models although their data can be used to build good models. I don't know whether they use SL data to build their own or just use SL models though.

For staff, I know of 3 main members: Max Erenberg, Matt Prefontaine, and Larry Simmons

Simmons is the AGM. I've had breif discussions as a potential vendor when HockeyData was potentially selling data to Jets. I wasn't part of all those convos (I mostly just deal with IP for the company) but I did had brief talks to Cheveldayoff, Simmons, and Maurice during that time. I'm guessing of the 3 of them he uses the data the most, but I doubt any of the grunt work. My guess is his main usage of analytics is Capologist type stuff but I'm throwing a dart in the dark without any real knowledge.

Prefontaine is the Video Coach/Analyst for the Jets and deals with analytics as well. I don't know what his specific role is, if maybe it's more traditional where he just links video to analytics or more hybrid modern role. I've never had any hockey convos with him. I did talk to him once about bringing up a young family and he seemed super nice though.

Erenberg is the Jets only designated hockey analytics person that I know of. I have never met or talked with him. Our only contact is that we both follow each other on Twitter. That's it. I do know of people (some of the ones I listed above) who have talked to him, and they all said he was nice. But again, I don't know his standing on any theoretical or philisophical standings on modeling and other stuff to make a call.

awesome insight

 

[voyageur]

I would be surprised to see them in the lineup. Save for a significant injury to someone like Morrissey or Pionk, I don't expect to see Heinola in the line-up.

He doesn't kill penalties, so Mo won't want him in a depth role.

Well killing penalties is important. It could be pointed out as the single biggest reason we lost to Calgary in the playins.

I noticed Samberg is getting top reps on the Moose, but curiously with Sauntner who seems to be on the ice for every goal against the Moose give up. Heinola actually has more PK minutes than Stanley on the Jets this year, so I don't know if he is getting the 2nd reps.

Forbort and De Melo aren't getting knocked out of the lineup because that's their role on the team. If we went with Morrissey-Samberg as the 2nd unit, well I'd say that there is room for Heinola, who I think will eventually impact our breakouts in a positive way.

But it's hard to argue that Poolman should come out of the lineup right now, so I'd say there is probably only one spot for contention, for now. And if it's Stanley to take, while Samberg and Heinola develop, don't see a problem with that.

I'd like to see the Moose get some chemistry developed between the young players. Send Ves and Gus down. Give Ves the PP reps that he can't get on the Jets, give Gus the PK reps he can't get on the Jets. Have Heinola QB a PP of Ves, Perfetti and who knows that chemistry might pay dividends down the road? That's part of development. Where players get familiarized with other players. Seemed to pay dividends with the 2015 line last year. I'd be looking at Samberg-Kovacevic as the top PK unit on the Moose, because both those 2 have a future with the Jets.

 

[Whileee]

I just think of it this way, Jets waiving Beaulieu or putting him in the pressbox and never letting him on the ice helps their on ice product by similar amount.

Keeping him on the roster though does help:
* expansion draft
* guy that some teammates like
* veteran presence bs

Oh, I agree with sitting Beaulieu instead of Stanley, Heinola or Samberg, but I don't think they'll waive him and instead keep him around as expansion draft fodder.

 

[Whileee]

You really want me to bring in receipts...

Okay here we go, guys that I know use RAPM or similar models and would suggest not using +/- for evidence of anything:
(CAR) - Eric Tulsky, Kevin Kan
(COL) - Arik Parnass, Dawson Sprigins (oh he brought RAPM to hockey)
(CBJ) - Josh Flyn
(CHI) - Stan Bowman
(FLA) - Cam Lawrence, Josh Weissbock
(LAK) - Rob Volman
(MIN) - Mat Sells
(NJD) - Matt Cane
(PIT) - Sam Ventura
(SEA) - Alexandra Mandrycky, Namita Nandakumar, Dani Chu
(TBL) - Michael Peterson
(TOR) - Kyle Dubas, Eric Joyce, Cam Charron, Daryl Metcalf
(VAN) - Jonathan Wall, Aidan Fox
(WSH) - Tim Barnes

That's just ones that I personally know that would say that RAPM and similar models have their utility and traditional +/- does not... there's more out there than that.
Some would point out that these models are imperfect (as would I, I just published a piece on Hockey-Graphs today about that), but at least the model answers questions. Traditional +/- just answers what was their +/-.

Agree...

Any NHL team that uses raw plus-minus and not various adjusted shot metrics and rational goal metrics isn't being "traditional", they are being obtuse. NHL teams have looked at "scoring chances" and other ways of measuring performance for a long time, but now there is evidence for the utility of various shot metrics that might be more reliable and robust. Smart teams will improve measurement and analytical techniques where possible. As Garret points out in his recent Hockey Graphs post (well worth a read), there is also room to look for more qualitative methods to improve the use and interpretation of the quantitative approaches.

So leave the plus minus and random facts to Sawyer and Beyak, but NHL teams need to focus on more useful metrics and models.

 

[Whileee]

I know about many non-WPG teams more than I know about WPG.

I know they do import data from SportsLogiq, which also comes with models. I'm not a huge fan of SL's models although their data can be used to build good models. I don't know whether they use SL data to build their own or just use SL models though.

For staff, I know of 3 main members: Max Erenberg, Matt Prefontaine, and Larry Simmons

Simmons is the AGM. I've had breif discussions as a potential vendor when HockeyData was potentially selling data to Jets. I wasn't part of all those convos (I mostly just deal with IP for the company) but I did had brief talks to Cheveldayoff, Simmons, and Maurice during that time. I'm guessing of the 3 of them he uses the data the most, but I doubt any of the grunt work. My guess is his main usage of analytics is Capologist type stuff but I'm throwing a dart in the dark without any real knowledge.

Prefontaine is the Video Coach/Analyst for the Jets and deals with analytics as well. I don't know what his specific role is, if maybe it's more traditional where he just links video to analytics or more hybrid modern role. I've never had any hockey convos with him. I did talk to him once about bringing up a young family and he seemed super nice though.

Erenberg is the Jets only designated hockey analytics person that I know of. I have never met or talked with him. Our only contact is that we both follow each other on Twitter. That's it. I do know of people (some of the ones I listed above) who have talked to him, and they all said he was nice. But again, I don't know his standing on any theoretical or philisophical standings on modeling and other stuff to make a call.

Jets and other NHL teams would be smart to connect with universities / colleges to plug into analytic talent there ( could probably get it cheap). Problem might be nondisclosure expectations, since academics tend to want open sharing of methods and peer review. At the very least, NHL teams would probably benefit from having neutral external reviews of their data / analytic operations to assess methods and processes. Otherwise, you can have wacky charlatans take over and steer a franchise into bad proprietary methods.

 

[voyageur]

Agree...

Any NHL team that uses raw plus-minus and not various adjusted shot metrics and rational goal metrics isn't being "traditional", they are being obtuse. NHL teams have looked at "scoring chances" and other ways of measuring performance for a long time, but now there is evidence for the utility of various shot metrics that might be more reliable and robust. Smart teams will improve measurement and analytical techniques where possible. As Garret points out in his recent Hockey Graphs post (well worth a read), there is also room to look for more qualitative methods to improve the use and interpretation of the quantitative approaches.

So leave the plus minus and random facts to Sawyer and Beyak, but NHL teams need to focus on more useful metrics and models.

I've heard Brian Burke talk about +/- on Sportsnet, Paul Maurice talk about it. It's not the primary stat to measure a player, but this shot based statistical approach hasn't been proven in any way. In fact a team like the Islanders destroyed that model in the playins. So there's something more to analyze. I'd like to see real analysis beyond shots, shots, shots. One on one battles, defensive stops, passing efficiency, which was something that the Liiga tracked, zone possession. Because inevitably with any team sport, you are never getting the whole picture, in a team full of variables, and opposing variables, identifying a player by a single stat. And analytics are still raw data, like Covid statistics, for instance. Which is not to slight G-man in any way because he provides a lot of useful information.