WHAT’S HE BUILDING IN THERE?

by Lowetide

Yesterday the NHL saw a fascinating bit of team building from the Florida Panthers and Arizona Coyotes. Basically, the Panthers made cap room for themselves and the Coyotes added the modern Ethan Moreau. I get the Arizona side—that club is collecting prospects the way I did with my Roto team 20 years ago (I had Vlad Guerrero and Andruw Jones when they were kids). The Florida side is the unusual item, and I want to explore that a little today.

Interesting. Let’s have a look at Rowe’s roster and see where he might want to spend some of that walking around money earned for him by trading Crouse.

PANTHERS CURRENT PROJECTED LINEUP

PANTHERS CURRENT

I like the Florida roster, but there are some weak points. I have posted these depth charts on the fly, but a 2LD, a 2RW and maybe an extra center would be on the shopping list. In terms of depth? Probably have room to move a RHD (Mark Pysyk?) or a prospect, but there is room for improvement in areas of this roster.

OILERS CURRENT PROJECTED LINEUP

OILERS PROJECTED ROSTER AUG 17

What is PC building in there? For me, areas of need include RHD and center (depending on how the team handles Leon and RW). Question: Is there a deal to be made between these two teams? Mark Pysyk? Derek McKenzie?

By the way, a lot of Edmonton’s depth problems up front go away in two simple moves: Keep Leon at center, play Yakupov on the McDavid line. It’s the right thing to do for this roster. I doubt it happens.

The giant blue has now played two KHL games (averaging 17:58 a night) and had three shots in yesterday’s game. No points for the big man yet, I would be interested to see if he is getting any power-play time. Early days, good to see him playing a regular shift.

We should find out soon where Matt Benning is going to sign and the Oilers are apparently in the mix for the young defender. If he does sign, Benning would represent the fourth college player signed by Chiarelli this summer (Drake Caggiula, Nick Ellis and Patrick Russell). That would represent almost 10 percent of the active pro rosters (Oilers and Condors) this coming season, and give credit to Chiarelli for overcoming a ghastly 2014 draft (in part) through these signings. There was a hole, but this trio (and Benning) would represent a pretty nice recovery.

Next summer, the Oilers will have a better graduating class. Connor McDavid is already in the NHL, and the team has signed Caleb Jones plus Ethan Bear already. Ziyat Paigin may come over a year from now and John Marino is developing well. As of this morning, Edmonton has signed one player from their 2014 draft (Leon) and have another (Lagesson) tracking well.

LOWDOWN WITH LOWETIDE

A busy Friday show, 10 this morning on TSN1260. Scheduled to appear:

  • Steve Lansky, BigMouthSports. How much do we miss the Olympics?
  • Corey Graham, TSN1260. Oil Kings camp gets underway!
  • Matt Iwanyk, TSN1260. Eskimos enter a vital month, and where will the Oilers finish?
  • Brian Saipe, Racecar driver. Brian has been going out to the track forever, we will preview his weekend.

10-1260 text, @Lowetide on twitter. See you on the radio!

119 comments
0

You may also like

0 0 votes
Article Rating
119 Comments
Newest
Oldest Most Voted
Inline Feedbacks
View all comments
Pouzar

theres oil in virginia: This is the best post I’ve ever read here.Of course, recency bias is a real thing, but this is a very good post. “VOR’s Lament”

+1

theres oil in virginia

VOR:
I would like to build on something GMoney said in a recent thread. He is concerned that there is an over reliance on linear regressions and that he would like to see more scatter plots.

He was responding to a question and indicating some doubts about a particular piece of research. However, he has expressed this opinion several times.

I am going to go a lot further. I think one of the factors that prevents hockey analytics being taken seriously is how the results of current work are being presented. I have always been a bit sceptical about the entire idea of possession metrics. That is because proponents of the theory/hypothesis struggle with basic statistical concepts.

First of all there is the aforementioned over reliance on linear regressions. I am never sure if the writers, who are by and large amateurs, don’t know that r and p values can be close to 1 or 0 respectively and still tell you precisely nothing. Or are they just lazy or hiding things. Whatever the case , it s a fatal flaw. To get taken seriously, for us as readers to really trust your work, you need a second ,level of analysis. In that level of analysis you need to present confidence intervals. Remainder plots should also be compulsory. A JB test would be a bonus.

It is entirely possible that many of the key papers about possession metrics are wrong. In fact, it is an even money bet. Geoff Cumming has shown that 40% to 60% of all papers presented with only r or p values are wrong. It is very easy for the researcher to fool themselves and see relationships that don’t exist. Yet that is all we ever get, if that.

There is also a complete failure to test for non linear regressions, which are actually far more common than linear relationships in the natural world, especially when exploring human activities and responses.The linear regression approach eliminates outliers, thus the need for remainder plots. These outliers may be important. For example, the age at which players scoring peaks is clearly a fat tail curve, probably a Pareto distribution. Simply put, a linear regression gives us a high r value around an age of 23.5. However, more players peak somewhere in the fat and long tail. In other words, more than half of all players will peak after 23.5.

Worse, and my pet peeve, is that the sample sizes are ludicrously small yet nobody is deploying the statistical tools that are used for dealing with small sample sizes. The best we get is an off hand comment about sample size issues. There are numerous fruitful approaches that could be used to enrich our understanding of the datasets and their reliability. They aren’t even difficult to learn to use.

Then typically, and all power to GMoney for trying to change this with video tutorials and other tools, the fact is that most of the research in hockey analytics is presented in a way that is not user friendly.

I am not here to criticize. Or at least not only to criticize. I am have been trying to figure out how to make things better. So here is my suggestion. We all agree to use Geoff Cumming’s new book, Introduction to the New Statistics, as our style guide. It strikes a beautiful balance between the seen
him good approach and that of the math nerds. By the way, Geoff assumes no familiarity with stats and teaches all the fundamental concepts starting from scratch.

My real reason for wanting to see the data analyzed more thoroughly is that we, and read the last few months of blog threads here if you don’t believe me, are devolving into a community of dichotomies. We see everything in terms of yes/no, black/white, and with a great and linear certainty we are right. That certainty breeds the contempt and arrogance that turns so many non mathletes against analytics.

Geoff sees the world in curves and shades of grey where he isn’t certain about anything. He has collected and refined a bunch of very old statistical concepts into a bold new package he calls The New Statistics. If every one working in hockey analytics would adopt it our knowledge of the game we love would take a quantum leap forward.

This is the best post I’ve ever read here. Of course, recency bias is a real thing, but this is a very good post. “VOR’s Lament”

JD_Wry

blainer: Man that is something I have not heard in a long time.

My Dad had a barrel when I was mid-teens. It wasn’t my first experience with the liquor, so when I skimmed some of it and tried it, I thought that my Dad had anticipated the move and spiked the barrel with gasoline.

VOR

I would like to build on something GMoney said in a recent thread. He is concerned that there is an over reliance on linear regressions and that he would like to see more scatter plots.

He was responding to a question and indicating some doubts about a particular piece of research. However, he has expressed this opinion several times.

I am going to go a lot further. I think one of the factors that prevents hockey analytics being taken seriously is how the results of current work are being presented. I have always been a bit sceptical about the entire idea of possession metrics. That is because proponents of the theory/hypothesis struggle with basic statistical concepts.

First of all there is the aforementioned over reliance on linear regressions. I am never sure if the writers, who are by and large amateurs, don’t know that r and p values can be close to 1 or 0 respectively and still tell you precisely nothing. Or are they just lazy or hiding things. Whatever the case , it s a fatal flaw. To get taken seriously, for us as readers to really trust your work, you need a second ,level of analysis. In that level of analysis you need to present confidence intervals. Remainder plots should also be compulsory. A JB test would be a bonus.

It is entirely possible that many of the key papers about possession metrics are wrong. In fact, it is an even money bet. Geoff Cumming has shown that 40% to 60% of all papers presented with only r or p values are wrong. It is very easy for the researcher to fool themselves and see relationships that don’t exist. Yet that is all we ever get, if that.

There is also a complete failure to test for non linear regressions, which are actually far more common than linear relationships in the natural world, especially when exploring human activities and responses.The linear regression approach eliminates outliers, thus the need for remainder plots. These outliers may be important. For example, the age at which players scoring peaks is clearly a fat tail curve, probably a Pareto distribution. Simply put, a linear regression gives us a high r value around an age of 23.5. However, more players peak somewhere in the fat and long tail. In other words, more than half of all players will peak after 23.5.

Worse, and my pet peeve, is that the sample sizes are ludicrously small yet nobody is deploying the statistical tools that are used for dealing with small sample sizes. The best we get is an off hand comment about sample size issues. There are numerous fruitful approaches that could be used to enrich our understanding of the datasets and their reliability. They aren’t even difficult to learn to use.

Then typically, and all power to GMoney for trying to change this with video tutorials and other tools, the fact is that most of the research in hockey analytics is presented in a way that is not user friendly.

I am not here to criticize. Or at least not only to criticize. I am have been trying to figure out how to make things better. So here is my suggestion. We all agree to use Geoff Cumming’s new book, Introduction to the New Statistics, as our style guide. It strikes a beautiful balance between the seen
him good approach and that of the math nerds. By the way, Geoff assumes no familiarity with stats and teaches all the fundamental concepts starting from scratch.

My real reason for wanting to see the data analyzed more thoroughly is that we, and read the last few months of blog threads here if you don’t believe me, are devolving into a community of dichotomies. We see everything in terms of yes/no, black/white, and with a great and linear certainty we are right. That certainty breeds the contempt and arrogance that turns so many non mathletes against analytics.

Geoff sees the world in curves and shades of grey where he isn’t certain about anything. He has collected and refined a bunch of very old statistical concepts into a bold new package he calls The New Statistics. If every one working in hockey analytics would adopt it our knowledge of the game we love would take a quantum leap forward.

OF17

blainer: Man that is something I have not heard in a long time. Damn near killed me as a teen. My brothers had a swish barrel made from rum.

I learned real quick to respect booze after that. Took me a full year before I could even drink again after a bad night of that stuff.

A lesson that I hope all you young readers never have to learn.

That’s a lesson that can only be learned by trial and error.

blainer

JDï™:
Centre of attention,

Buy some old rye barrels and make some swish. So much swish.

Nuge is on the U24 team as well.

Man that is something I have not heard in a long time. Damn near killed me as a teen. My brothers had a swish barrel made from rum.

I learned real quick to respect booze after that. Took me a full year before I could even drink again after a bad night of that stuff.

A lesson that I hope all you young readers never have to learn.

Mr DeBakey

HT Joe:
Hey Frank the Dog… why was Petry labelled as soft, but Eberle and RNH (who aren’t destructo machines) never get booed?

(and I’m not encouraging that we start… I’m just not sure why Petry or Hemsky were ever targets)

Petry is a soft d-man, like Gilbert before him. Ever see either crush an opposing forward with a bodycheck as they crossed the blueline? Or, drop the flippers? I thought not. Soft.
We hates soft D-men, yes we does.

Also, in which prairie province do you find Pardubice?
Also, Hemsky refused to talk to the media.
Also, he left the ice early after practice.

HT Joe

Hey Frank the Dog… why was Petry labelled as soft, but Eberle and RNH (who aren’t destructo machines) never get booed?

(and I’m not encouraging that we start… I’m just not sure why Petry or Hemsky were ever targets)

HT Joe

Frank the dog: I said the fans ran Petry out of town. I’m saying that Petry was a really good, home grown,RHD that was booed at home games and insulted aplenty on the various beer & cheer blogs around the place. The common cat call was that he was “soft”. As opposed to ….. Gritensity?

Some losers booing the player still shouldn’t ever dictate a trade. If it’s really that bad, why didn’t MacT just once tell the media what a great player Petry was (well, you know, before he traded them)?

I’m looking at my comment again, and it probably came off as rude… not my intent and sorry if that’s how it looked. Thanks for the decent response.

AND I think we may actually be agreeing 😀

Centre of attention

JDï™: He’s also got Scheifele and Couturier for PKing duties. It could be that Nuge eats kale chips from the press box for most of the tourney, really.

Lol. it would be so Oilers for Todd and peter to wait until the second round of names to nominate Nuge, then just press box him.

I’d put Schiefle on RW maybe? He can definitely rip it.

DBO

We need to go back to thinking pairs. I like the idea of a “kid” or defensively inept with 2 solid players.

Lucic-McDavid
Nuge-Eberle
Pouliot-Draisatl

Add where needed. Yak with McDavid. Maroon with Nuge. Puljarvi with Drai. I think that gives a solid 1B line in Nuge’s line. And Draisatls line will be solid with all 3 being solid two way players (even with their youth) and McDavid and Lucic can drag Yak along.

JD_Wry

Centre of attention: Nuge on the pk

He’s also got Scheifele and Couturier for PKing duties. It could be that Nuge eats kale chips from the press box for most of the tourney, really.

Edit: Boring Monahan on the PK! Just for you, Cameroon!

JD_Wry
Centre of attention

JDï™:
Centre of attention,


Group B
Team Finland
Team North America
Team Russia
Team Sweden

If they don’t finish in the top two spots of this group, McDavid and Nuge will never meet Weber. Draisaitl probably won’t PK for team Euro, and he’s got big bones!

Edit: And if McLellan isn’t running Eichel and Matthews ragged on the PK, he’s not doing it right!

He’ll probably run Nuge on the pk, because Nuge is fantastic at it.

Until he takes a slapshot to the knuckles.

Centre of attention

Bear scores for seattles team blue during scrimmage, nifty fake before sliding it under the goalie

Water Fire

Centre of attention: If McDavid gets hurt during the world cup, what will we do with ourselves?

Katz owes us all, at least Lowetidians, a minimum one week cheap but decent all in holiday. My line in the sand for the abuse I’ve/we’ve taken from them.

JD_Wry

Centre of attention,

Group B
Team Finland
Team North America
Team Russia
Team Sweden

If they don’t finish in the top two spots of this group, McDavid and Nuge will never meet Weber. Draisaitl probably won’t PK for team Euro, and he’s got big bones!

Edit: And if McLellan isn’t running Eichel and Matthews ragged on the PK, he’s not doing it right!

Frank the dog

HT Joe: Nope… the fans did not trade away Petry… MacT did.That failure is on MacT (and Katz for empowering MacT).
(1) – If you’re going to claim that MacT couldn’t re-sign Petry because Petry hated the coach, that’s on MacT for hiring said coach.
(2) – If you’re going to claim that MacT couldnt re-sign Petry because Petry was disgruntled in Edmonton, I’d propose that MacT giving several lesser blue liners more money probably contributed to said disgruntlement.
(3) – If you’re going to claim that MacT couldn’t re-sign Petry only because the fans didn’t give him enough love, you have even less evidence than I do with point #2.

Booo MacT!Boo!

I said the fans ran Petry out of town. I’m saying that Petry was a really good, home grown,RHD that was booed at home games and insulted aplenty on the various beer & cheer blogs around the place. The common cat call was that he was “soft”. As opposed to ….. Gritensity?

Centre of attention

JDï™:
Centre of attention,

Buy some old rye barrels and make some swish. So much swish.

Nuge is on the U24 team as well.

I’m down for swish.

Nuge, Drai, Sekera, AND McDavid are all at the world cup. All are extremely important to the upcoming season, what are the odds at least one of them blocks a shot and breaks a foot?

Our top 3 centers and #3 defenseman are at the mercy of Shea Weber’s slap shot.

Can we just not send our guys and they can name some ridiculous replacements we can all laugh at? Wouldn’t that be so much more fun, lol?

Pouzar

JDï™:
Centre of attention,

Buy some old rye barrels and make some swish. So much swish.

Nuge is on the U24 team as well.

I hate World Cups.

JD_Wry

Centre of attention,

Buy some old rye barrels and make some swish. So much swish.

Nuge is on the U24 team as well.

Centre of attention

JDï™: Yeah their preseason starts a little later than the Oilers, so that kind of makes sense, while still less than ideal.

Another 4 win October, and the gales of November will come early.

If McDavid gets hurt during the world cup, what will we do with ourselves?

JD_Wry

Centre of attention: Gerry Fleming

Yeah their preseason starts a little later than the Oilers, so that kind of makes sense, while still less than ideal.

Another 4 win October, and the gales of November will come early.

Centre of attention

JDï™: Woodcroft is McLellan’s assistant at the worlds as well, and it’s unknown how much time Chiarelli will need to spend as GM of the team while the tourney is on.

I’m cheering for the NAYS, but also kind of want them to get skidded as early as possible.

Until then, I don’t think anyone has mentioned who will be running TC. Bucky and MacT?

Gerry Fleming and Ian Herbers I think? I’m sure Bucky and Keith Gretzky will be hanging around too.

Peter Chiarelli will be in Toronto during the World Cup, along with a bunch of other Gm’s. I’m sure that is fertile ground for trade discussion FWIW.

JD_Wry

Centre of attention: Todd as well as Drai, Nuge, Sekera, and McDavid all will miss varying amounts of training camp.

Woodcroft is McLellan’s assistant at the worlds as well, and it’s unknown how much time Chiarelli will need to spend as GM of the team while the tourney is on.

I’m cheering for the NAYS, but also kind of want them to get skidded as early as possible.

Until then, I don’t think anyone has mentioned who will be running TC. Bucky and MacT?

Woodguy

Caramel Batman:
Frank the dog,

The fanbase is positive on Chiarelli.Proof positive is that they like the trade simply because Chiarelli made it.Look at the Staples link above.There is a poll on Chiarelli’s performance and gave him a B.

The guy is teflon, completely immune from criticism by either the local media or fanbase.The only people who dislike him are anonymous folks on the internet.

What colour is the sky in your world Verdad?

Centre of attention

JDï™: Especially given who will be running it.

Yeah i just remembered that Todd as well as Drai, Nuge, Sekera, and McDavid all will miss varying amounts of training camp.

If team NA goes far like I think they will, Todd won’t be back until almost October lol. At least they will be playing a high level of hockey so it won’t be like they are rusty.

*edit* I swear to the gords if any of those 4 guys get hurt in the World Cup I’m hoping no Oilers are ever “good enough” to get selected again! Hold them back from it with pretend “injuries” to make sure they don’t get real ones damnit!

Centre of attention

Lowetide: Agreed, If the Oilers use Maroon as a Pouliot proxy, it will not work. Jmo.

I think Pouliot has plenty of value on the third line mentoring the yutes. He’s the closest thing to a Pisani we got.

Its not a hard target either way like I said above. I suspect we are both right, at different times of the season. The lines will be written in pencil for at least the first couple months.

JD_Wry

Centre of attention: Training camp is going to be so very interesting.

Especially given who will be running it.

Centre of attention

Lowetide: Oh, I am certain people disagree, and I do think he could score 20. I mentioned how in my RE
https://lowetide.ca/2016/08/04/re-16-17-patrick-maroon-blue-collar/

You mentioned that he could, *if* Todd goes against your RE and plays Pouliot on the 3rd line like I keep saying 😛

Training camp is going to be so very interesting.

Water Fire

I doubt that the Oilers won the Lucic sweepstakes by telling him he’d get all the Nuge he could stand.

Really he was hired to ride shotgun at least for a bit unless there is no chemistry, but the e know from those who looked into it before the whole analytics blogosphere went pro that chemistry really means good enough linemates. Good players can play with other good players.

And its a good idea IMO to have deterrent right there that will deter with prejudice. Lucic and McDavid can carry Yak so I would put Eberle with Nuge and rotate everyone else based on opponent and health.

Centre of attention

Lowetide: No. I have Maroon on the 3line. Always have.

Yeah I bet he spends time on a variety of lines like I said. It would be hard for me to see him scoring 20 with exclusively 3rd line minutes though. You have said in your summaries and I quote “Patrick Maroon: People are betting low on him. I think he could score 20.” I think he would need at least a couple of McDavid minutes to hit that kind of career high. Maybe they come on the power play?

Centre of attention

Lowetide: It would be better, but for me Pouliot should be with Nuge or McDavid. He is a very good LW.

You keep telling me in your TC summarys that Maroon might score 20. If he’s scoring 20, he’s in the top 6. Lucic is definitely in the top 6 at that price point, hell or high water. So that means Pouliot is on the third line right?

I bet its a moving target for most of the year and both of our rosters see the ice for multiple games, as well as some combination we would never think of. (Pouliot 3C anyone? Hahahaha…*tugs collar nervously*)

HT Joe

Frank the dog:
I sometimes think this fan base is the Oiler’s second biggest problem. They run players out of town because they are not mean enough and are therefore considered soft: Hemsky, Petry, for example.

Nope… the fans did not trade away Petry… MacT did. That failure is on MacT (and Katz for empowering MacT).
(1) – If you’re going to claim that MacT couldn’t re-sign Petry because Petry hated the coach, that’s on MacT for hiring said coach.
(2) – If you’re going to claim that MacT couldnt re-sign Petry because Petry was disgruntled in Edmonton, I’d propose that MacT giving several lesser blue liners more money probably contributed to said disgruntlement.
(3) – If you’re going to claim that MacT couldn’t re-sign Petry only because the fans didn’t give him enough love, you have even less evidence than I do with point #2. 🙂

Booo MacT! Boo!

Centre of attention

Lowetide: I would not play Leon and Nail, not this season.Thats what putting 10 with McDavid makes so much sense. Share the talent, share the risk.

What if Pouliot is on the other wing? He can rain-delay when Yak has hiccups and Drai gets tired.

Centre of attention

Lowetide,

What do you think of the line combos I posted above, LT?

Does that leave Nuge with too little re-inforcements? I think it is somewhat balanced, and I would love to see that Pou-Drai-Yak line get a crack at the soft underbelly of teams while we are at home.

JD_Wry
Chachi

Lowetide: This is wildly incorrect.

Yes, I hear his family is not wild about him.

Chachi

Cameron: ‘Make the Oilers Great Again’

“Best of luck to the Flames this season”

JD_Wry
Cassandra

Frank the dog,

The fanbase is positive on Chiarelli. Proof positive is that they like the trade simply because Chiarelli made it. Look at the Staples link above. There is a poll on Chiarelli’s performance and gave him a B.

The guy is teflon, completely immune from criticism by either the local media or fanbase. The only people who dislike him are anonymous folks on the internet.

Ca$h-McMoney!

Caramel Batman: Chiarelli is worse than

Chiarelli, on the other hand, traded away one of the best players of his generation for a second pairing defenseman.

I like Hall, I really do.

One of the best players of his generation?

51 points in 67 games is his pace the last 2 years. This assumption that the “real Hall” is a 100 point dynamo he was starting to maybe look like 3 years (and plenty of injuries) ago just blows my mind.

He’s really good. Better than Larsson, equalizing for position. But man, did your comment over play your hand.

Frank the dog

I sometimes think this fan base is the Oiler’s second biggest problem. They run players out of town because they are not mean enough and are therefore considered soft: Hemsky, Petry, for example. They are down on Chia for decisions made which they don’t understand because they are not proxy to the facts driving the decisions. Hall is a shining example. How many people have reduced esteem for Chia because he made what can only be called a bad trade if the parameters are limited to statistical on-ice performance and the two players directly involved in the trade.
That’s one wedge of the pie, folks. We can all agree with what we can see but we cant argue that what we don’t know or see doesn’t exist. That is foolishness. The whole is greater than the sum of the parts. The whole team is more balanced after the Hall-Larsson swap than before, not in and of itself by any means but rather as a part of everything Chia has done and will do from the end of last season to the end of next season.
If a trade involved getting Larsson and Lucic for Hall, some would complain but not as many as see Hall vs Larrsson as a closed circuit.
From what I have heard and seen Hall was totally blindsided. Hall had no intention of leaving the Oilers. Hall is a alpha male. Hall did not play well with CMD but Hall did drive a 1A line. with Drai. and Nuge also ran a 1A line with Poo and Ebs. The statistical unicorn of 3 scoring lines was swapped out for balance. Dare I say, I would not be surprised if it always was Hall tha tChia wanted to trade, and that we will see if this was the case by the team’s cohesiveness this season compared to last. I also think the psychos up and down the lineup and behind the blue (hello Darnell) should keep the opposing teams from running our skill players from every angle even if there are some injuries.

Cameron

Chachi: Scott Baio is obviously a gigantic idiot.

‘Make the Oilers Great Again’