THE SIXTH ROUND (2016 EDITION)

by Lowetide

For the Edmonton Oilers, the sixth round is where luck landed in 2010 and Brandon Davidson is the prize. Make no mistake, if they knew he was going to be this player, Steve Tambellini and Stu MacGregor would have called his names several rounds previously.

There’s a 15.8% chance a player chosen in this round plays 100 NHL games—which sounds like ‘why bother?’ but some exceptional players were taken this late. Among Oiler draft picks 2000+ who have played 100 games, we can count no one. The last Oilers selection from the sixth round to play 100 games? Anatoli Semenov, 1989. It’s been awhile, but Davidson should get there this coming year.

I’m hopeful we can have a rational conversation about the Edmonton Oilers in the sixth round 2008-15. No grand statements on either side, no claims of victory, just an honest assessment of the way things are, the good and bad arrows, and a discussion in regard to useful players coming out of this area of the draft. In order to do this, we absolutely have to agree on some things:

  • We can’t judge a drafted player weeks after his selection.
  • We can count arrows, and they are a good indicator.
  • We cannot expect all of the picks to succeed, that isn’t reasonable.
  • We have come to some kind of conclusion about a line in the sand and agree to it (the Cullen grades)

OILERS SIXTH-ROUND PICKS, 2008-16

  • 2008: Teemu Hartikainen (Kevin Lowe) (Stu MacGregor)
  • 2010: Brandon Davidson (Steve Tambellini) (Stu MacGregor)
  • 2010: Drew Czerwonka (Steve Tambellini) (Stu MacGregor)
  • 2012: John McCarron (Steve Tambellini) (Stu MacGregor)
  • 2013: Ben Betker (Craig MacTavish) (Stu MacGregor)
  • 2014: Tyler Vesel (Craig MacTavish) (Stu MacGregor)
  • 2015: John Marino (Peter Chiarelli) (Bob Green)
  • 2016: Aapeli Rasanen (Peter Chiarelli) (Bob Green)
Teemu Hartikainen and Brandon Davidson made the NHL, but the newer names like Betker, Vesel and Marino are miles away—and may never get close to playing in the show. The last two names chosen have some nice things to tell us, more on that in a minute.

2015 SUMMER (NHL GAMES PLAYED)

  1. Brandon Davidson 63
  2. Teemu Hartikainen 52

Davidson has a long way to go if he is going to catch the top sixth-round pick in franchise history—Steve Smith played in 804 games—but he is a welcome asset all the same.

ARROWS

  • Productive in Europe/AHL: Teemu Hartikainen
  • “It’s Over” arrows: Drew Czerwonka, John McCarron
  • Bad arrows:
  • Lukewarm arrows: Ben Betker, Tyler Vesel
  • Good arrows: John Marino
  • Very good arrows: Brandon Davidson
  • New arrows: Aapeli Rasanen

ONE BY ONE

  • L Teemu Hartikainen: Hartikainen is now firmly entrenched as a solid KHL forward, having posted three good years in a row. His NHLE from last season (47 points) makes me wonder if he might get an NHL shot. He is 26 and there is size and skill.
  • D Brandon Davidson: Even last year at this time, there was no certainty about an NHL career. Edmonton was graduating Darnell Nurse and had traded for Griffin Reinhart—but Davidson trumped both and all emerging blue with a fine year and won a two-year deal.
  • L Drew Czerwonka: As happens so often with prospects, he was injured often during junior and passed on a pro career due to the rigors of the game.
  • R John McCarron: Edmonton did not sign him a year ago, McCarron catching on with Scranton/Wilkes-Barre Penguins. He ended up flourishing in the ECHL with the beautifully names Wheeling Nailers, and I would expect more AHL time this coming season.
  • D Ben Betker: Giant blue spent most of his first pro season in the ECHL. Blocked by the Leftorium, he probably benefited from the year at a lower level, and we could see him take a regular shift in Bakersfield in 2015-16.
  • F Tyler Vesel: Showed promise in his first NCAA season and came close to the same production in year two. A two-way forward, he is halfway through his college career and worth following—although a pro contract is far from guaranteed over the next two summers.
  • D John Marino: Playmaking defender with size, good skater, nice shot. He stepped up to USHL level and maintained his offense, a nice arrow. This coming season he will be in the NCAA at Harvard, and that range of skills will receive another test.
  • C Aapeli Rasanen: Surprised he fell this far, the scouting reports and the numbers suggest this is a player of interest for Oilers fans. Played in Jr. Sm-Liiga last year and will suit up for the Sioux City Musketeers of the USHL this season.

 

117 comments
0

You may also like

0 0 vote
Article Rating
117 Comments
Newest
Oldest Most Voted
Inline Feedbacks
View all comments
Riley Reid

Hi LT and all,

First time poster, long time lurker. Thanks for all the level-headed discussion over the years and the entertainment, of course. Wonderful corner of the Al Gore, this is.

I’ve never felt compelted to post on any blog let alone this one, but I was watching a movie trailer on the intertubes and saw something that caught my eye at the end of the trailer for “Now You See Me 2″…

Intrigued, I delved further in to it: http://www.imdb.com/name/nm1095804/?ref_=fn_al_nm_1

Maybe I missed it, but not sure if anyone else has commented on how cool this is before. I am officially even more happy that this guy is our Oil’s GM!

Probably old news for all of you, but glad I found that little tidbit.

Thanks all. And Go Oil, of course.

RR

Andy Dufresne

Perspective. I’m Canadian. My concern about the wall (walls) is that I like to go to Mexico every other year. If they start blocking entry based on things like perception/color/character I might not get in!!
Loves me some Mexico. Beautiful country. Warm and Friendly people. Never met a Mayan I didn’t like.

Woodguy: Thanks Donald

Andy Dufresne

Woodguy

Hawkings/Roman Empire retorts = Early Morning Nasal Coffee Enema for me

I was asking you yesterday about Qual Comp and how it’s created. Didn’t get a chance to thank you for your answer. And today BAM….the curtain gets drawn back to reveal some of the magic.

If Analytics is in the 3rd inning of a nine inning ball game, BIG thanks to Gmoney and You and LT( and a few others here) for allowing me/us the privilege akin to being a fly on the wall in Steve Wozniaks garage in 1976.

Keep up the good work and the generous attitude towards sharing.
Open Source baby Open Source!

(edit: So what you’re saying is that Gmoney is Woz and you are Jobs. Watch out G we all know how that one ended 🙂 )

Woodguy: Yeah, my contribution consists of:

“Hey G, think you could do this?”

“You need monster computing power to pull this off?Here’s some money to build your monster so I can feel like I contributed”

“Are you done yet?”

Cyanide

Woodguy,

“Everyone takes Kane over Eberle, but that doesn’t mean Eberle isn’t among the best RW in the league.”

I wasn’t bashing Eberle at all. I realize he is one of the best RW in the league.
And I also realize Kane has some warts in his game, but may very well be the best RW in league.
I am also open to the Idea of someone who was elite last yr, may not be elite next yr.

I agree with everything you are saying. Just wanted clarity on Kane, was confused somewhat of his omission from the list.

Woodguy

Cyanide:
Woodguy,
“We’ll have to make exceptions for players who are on the very far right end of the curve on a few metrics, but fail in another I think.”

That’s the only reason I asked. Kane is obviously an elite player.
For sure more elite than most players on that list.
I understand the metrics for the most part.
I just look at the list and see Jordan Eberle, who in my opinion, is comparable to a Patrick Kane (style of player, obviously not quite as talented)
But I feel like Eberle’s other metrics get zoomed a bit by his river pushing linemates.

I love Jordan Eberle, but if I am choosing a RW its Patrick Kane.

Edit: I am not bashing G’s and your work. Its excellent work. Just trying to understand it more myself.

We are trying to set parameters that will catch all the good players and exclude the not-as-good

It needs to pass the eye-test/sniff-test but you also have to be open to the idea that a player is better or worse than you think he is.

Everyone takes Kane over Eberle, but that doesn’t mean Eberle isn’t among the best RW in the league.

Woodguy

frjohnk: Well GMoney is like EBERLE.

His hands in the slot (keyboard) are top notch while his offensive IQ ( computer programming) is elite.

But terrible on the backcheck and needs to be pushed.

That’s where you come in.

Ha!

I’d be like Hitchcock to his Tarasenko and just make the wrong decision most of the time.

Woodguy

G Money,

Top Comp TOI 46577 CF 42.3 DFF 44.5
Muddle TOI 24173 CF 45.8 DFF 53.2
Dregs TOI 14914 CF 48.9 DFF 51.8

When I was looking at Larsson, my two basic conclusions were:

1) Shots For go to die when he’s on the ice.

2) He’s an elite shot suppressor as he gets, by far, the hardest DZFO% in the NHL AND manages to get a -RelCA/60 AND -RelxGA/60 AND +RelGF% while doing it (very few pull this off)

So, I think a metric where Larsson will jump off the page (*crosses fingers*) is RelDFA/60 against the best.

Also,

We don’t know IF ANY Dmen post a +50% DFF against the best yet.

There are probably some.

I can’t wait to look at it!!!!

frjohnk

Woodguy: Yeah, my contribution consists of:

“Hey G, think you could do this?”

“You need monster computing power to pull this off?Here’s some money to build your monster so I can feel like I contributed”

“Are you done yet?”

Well GMoney is like EBERLE.

His hands in the slot (keyboard) are top notch while his offensive IQ ( computer programming) is elite.

But terrible on the backcheck and needs to be pushed.

That’s where you come in.

Cyanide

Woodguy,
“We’ll have to make exceptions for players who are on the very far right end of the curve on a few metrics, but fail in another I think.”

That’s the only reason I asked. Kane is obviously an elite player.
For sure more elite than most players on that list.
I understand the metrics for the most part.
I just look at the list and see Jordan Eberle, who in my opinion, is comparable to a Patrick Kane (style of player, obviously not quite as talented)
But I feel like Eberle’s other metrics get zoomed a bit by his river pushing linemates.

I love Jordan Eberle, but if I am choosing a RW its Patrick Kane.

Edit: I am not bashing G’s and your work. Its excellent work. Just trying to understand it more myself.

Woodguy

frjohnk:
Woodguy,

Great job in all of this stuff.

*Does best Wayne impression when he first meets Alice Cooper*

Yeah, my contribution consists of:

“Hey G, think you could do this?”

“You need monster computing power to pull this off? Here’s some money to build your monster so I can feel like I contributed”

“Are you done yet?”

Woodguy

G Money,

Also,

I love you long time brother!!!

WOOOOOOOOTTTTTTTTTTTTTTTTT!!!!!!!!!!!!!!!!

Also,

5. And lastly, remove any forwards whose time on ice per game is less than the average time on ice for all forwards from step #2. If the coach distrusts that player to that extent, we should assume he’s not elite. n=95

Not sure about this one.

Were there some weird results you wanted to filter out?

I’m worried that the metric *might* eliminate some emerging young players

Then again, we’ll have their results to see when we run analysis.

Is Mr. Haul set up to do team metrics yet so we can look at Rel?

Also,

I love you long time!!

frjohnk

Woodguy,

Great job in all of this stuff.

*Does best Wayne impression when he first meets Alice Cooper*

Woodguy

G Money,

Top Comp TOI 46577 CF 42.3 DFF 44.5
Muddle TOI 24173 CF 45.8 DFF 53.2
Dregs TOI 14914 CF 48.9 DFF 51.8

Remember that NJD is a shitty team and that DFF is *probably* a positive RelDFF vs top comp (I’m hoping really), but still sobering stuff.

Put’s Hall/McDavid into perspective too, as their RelDFF vs top comp is probably +10 or so.

frjohnk

G Money: McD
Top Comp TOI 18339 CF 47.6 DFF 51.8
Muddle TOI 13593 CF 53.2 DFF 51.9
Dregs TOI 6090 61.2 DFF 69.7

Hall
Top Comp TOI 37878 CF 48.2 DFF 49.0
Muddle TOI 23785 CF 52.4 DFF 54.6
Dregs TOI 12713 CF 56.5 DFF 60.1

Sekera
Top Comp TOI 39771 CF 45.5 DFF 45.8
Muddle TOI 25021 CF 49.6 DFF 48.8
Dregs TOI 15739 CF 55.3 DFF 55.9

Gryba
Top Comp TOI 20108 CF 47.4 DFF 48.0
Muddle TOI 16326 CF 49.8 DFF 45.5
Dregs TOI 12256 CF 52.7 DFF 50.5

Nurse
Top Comp TOI 30354 CF 41.2 DFF 41.6
Muddle TOI 25345 CF 48.3 DFF 48.9
Dregs TOI 15341 CF 51.1 DFF 48.4

Korpikoski
Top Comp TOI 16990 CF 37.3 DFF 40.1
Muddle TOI 16510 CF 46.6 DFF 44.2
Dregs TOI 12661 CF 47.5 DFF 41.9

Adam Larsson:
Top Comp TOI 46577 CF 42.3 DFF 44.5
Muddle TOI 24173 CF 45.8 DFF 53.2
Dregs TOI 14914 CF 48.9 DFF 51.8

In this list we have a collection of guys who are close to being the best player in the game in McDavid, to good hockey players in Sekera all the way to bottom of the league players in Korpel Punishment

All play against the toughs, the middling players and the softs.
The only thing that is different is the allocation of ice time of each player against each grouping.
Thats why it is so important to have actual NHL players in the lineup, top to bottom. 3rd pairing Dmen, 3rd and 4th line forwards will at times during the course of the game will be up against the toughs. If your bottom of the roster players can limit the damage playing against the toughs, you most likely have a good team.

For far too long we had too many tweeners and I would imagine they got their heads kicked in big time against all 3 groupings.

I would say a tweener is a guy that gets his head kicked in by all 3 groupings like Korpel Punishment.

I had always felt that Nurse did OK as a rookie when not playing the toughs but I could never prove it with numbers. Now I can. Thank you.

Man this is impressive stuff.

Woodguy

Cyanide:
G Money,

No Patty Kane? am I blind? You have his line mate Panarin, but not him?

Probably because he didn’t have a +RelCor

Kane has never been a river pusher/good possession player

Best scorer in the league for a while now though.

We’ll have to make exceptions for players who are on the very far right end of the curve on a few metrics, but fail in another I think.

We’ll look at it.

Kane is elite, but in very unique way (scores a ton, pretty meh corsi)

Woodguy

G Money,

(which is part of the effect that confounds the current QoC metrics, and which we hope to mitigate, though we’ll never get rid of it as it represents violence inherent in the system).

Help! Help! I’m being repressed!

Woodguy

haters:
Omg wg I was just bugging man. No disrespect. I did read it. I read all your shit… Was bieng glib and silly.

In that case I apologize.

It did not read that way.

Sorry

frjohnk

G Money:
First thing to note here is you’ll see that a lot of players appear to be playing a lot against top comp.Part of that is because this is still a test version using a pretty lax definition of ‘top comp’.As that gets tightened down (WG and I intuitively think the list should be sixty players or less – average of about two per team), those numbers will shift away.

Some of it will inevitably remain, though, precisely because the best players tend to get the most time on ice, so they show up a lot (which is part of the effect that confounds the current QoC metrics, and which we hope to mitigate, though we’ll never get rid of it as it represents violence inherent in the system).

McD
Top Comp TOI 18339 CF 47.6 DFF 51.8
Muddle TOI 13593 CF 53.2 DFF 51.9
Dregs TOI 6090 61.2 DFF 69.7

McDavid faces a decent bit of the best competition already, though the gap you see here is actually quite modest.I’d call it ‘modestly sheltered’.Certainly, he rarely sees the easy competition.He’s mostly holding his own (47.6% CF and 51.8% DFF) even then – he’s a dangerous player no matter who is on the ice against him.

And he eats the dregs for breakfast, and spits out their bones.

Hall
Top Comp TOI 37878 CF 48.2 DFF 49.0
Muddle TOI 23785 CF 52.4 DFF 54.6
Dregs TOI 12713 CF 56.5 DFF 60.1

No surprises – Hall faces the best of the best night after night.And he holds his own (imagine what that number would look like with a consistently good D behind him).And like McDavid, he crushes the weak.

Sekera
Top Comp TOI 39771 CF 45.5 DFF 45.8
Muddle TOI 25021 CF 49.6 DFF 48.8
Dregs TOI 15739 CF 55.3 DFF 55.9

The TOI is as expected – like Hall, Sekera is fed to the wolves every night.Unlike Hall, he’s not holding his own, though.Perhaps a better partner, or easier competition.

Speaking of which – Sekera does just fine against the muddy middle, as expected.And manhandles the bottom of the roster.

Gryba
Top Comp TOI 20108 CF 47.4 DFF 48.0
Muddle TOI 16326 CF 49.8 DFF 45.5
Dregs TOI 12256 CF 52.7 DFF 50.5

This one does surprise me.Gryba faced more time against top comp (probably when he was with Sekera because of the righty shortage), and also does much better than expected against the ruffians of the world. The middle and dregs numbers make more sense to me.

A serviceable bottom pairing guy.

Nurse
Top Comp TOI 30354 CF 41.2 DFF 41.6
Muddle TOI 25345 CF 48.3 DFF 48.9
Dregs TOI 15341 CF 51.1 DFF 48.4

The Darnell Nurse story is writ large in these numbers.He’s spending too much time against top comp (the three numbers should be much closer together, like 26, 25, 20 instead of 30, 25, 15), and he’s getting eaten alive.

I guess a bit of good news is he’s muddling through against the muddle, and (mostly) holding his own on the bottom pairing.Next year is an important development year – a Top 4 D should be showing signs of being able to handle top comp by this point.

Korpikoski – I had to throw this one in.These numbers are about as sheltered as you can get, with Top and Middle neck and neck, with Dregs (the 10min TOI guys) modestly behind.

But it doesn’t matter who he faces, he generally gets crushed.And the WOWYs show he drags whoever he plays with into the mud with him.

Top Comp TOI 16990 CF 37.3 DFF 40.1
Muddle TOI 16510 CF 46.6 DFF 44.2
Dregs TOI 12661 CF 47.5 DFF 41.9

Addition by subtraction.

And last but certainly not least, Adam Larsson:

Top Comp TOI 46577 CF 42.3 DFF 44.5
Muddle TOI 24173 CF 45.8 DFF 53.2
Dregs TOI 14914 CF 48.9 DFF 51.8

Yes, he plays a crazy amount of top comp.Even moreso than Hall or Sekera, by quite a wide margin.

He does not do well – so I remain concerned that asking him to play top pairing here will require a pretty good developmental jump forward.We’d all be happier if he was slated at second pairing.

I am encouraged by his DFF numbers always being better than his CF numbers – the margin is wide enough that it doesn’t strike me as just random variation.Rather, to me it indicates that his reputation is earned, that he’s good at reducing the really dangerous stuff against.

And maybe with better forwards in front of him (except river pusher *sniff* Hall *sniff*), the ‘for’ side will go up and his numbers will show that much better.

I hope for the best but … I remain concerned.

ABSOLUTELY FANTASTIC!

Not only do you have a better quality of competition but also a better quality of teammate metric. Both are important
Love this road you and WG are going down.
Cant wait for more stuff.

magisterrex

Am I wrong or do I see ALL the Oilers’ bottom six in the “dregs” category?

Yikes!

rich

G Money,

Thank you for doing this. Wow is the best word to describe. Especially the comps on Nurse and Gryba.

The only forward that I was surprised by where he showed up was Versteeg. We’re very fortunate to have you and Asiaoil posting lucid thoughts in the middle of the night.

magisterrex

G Money,

This is awesome. Full stop.

Oil2Oilers

G Money,

Outstanding work, yet again you are increasing our understanding of the game. Thank you

spoiler

G Money: omissions?
I’m sure there’s a few, but I can tell you, compare this list to the top of the traditional QoC list, and it is night and day.

Stastny beating out Steen, is the one that leaps out at me. Pretty sure Steen is Hitch’s go-to guy.

Not sure Hudler is a top line player, but he’s definitely been kicking who he’s playing. I’ve wanted to sign him since March. Figgered it would be Nuge who was gone.

And agreed. Good list. Well done, bro!

spoiler

G Money,

I’m fucking thrilled. QoC is the measure I’ve been most unhappy with and it stretches back years.

Cyanide

G Money,

No Patty Kane? am I blind? You have his line mate Panarin, but not him?

G Money

Well, damn, it’s 3:30 but I finally have all this massive heaping of code running pretty much flawlessly over a MASSIVE new database, slick as slobber.

(Calculating the “vs band” TOI and shot metrics above takes about a half second per game, or 40 seconds per player). Running on an AMD A10 with 32 GB of RAM and the database on an SSD.

Wheeeeeeeee!

Good night, I will check back in tomorrow if there are any questions or comments.

Still lots of work to do, but DAMN ITS GOOD TO BE A (NERD) GANGSTA.

G Money

Philosophil,

I suspect he scores too much to be part of the dregs. His CFRel this year probably also kept him out.

Contrary to popular opinion, his shot metrics were decent this year – particularly impressive given how much time he spent with Korpi, who dragged Hendricks, Letestu, and Pakarinen to the bottom of the toilet.

His good numbers with McDavid were no fluke, and I believe neither was the fact that the other top line players scored nothing with Letestu.

There is far too much shitting on the Yak I would say.

G Money

Water Fire,

Ha ha, like with Granlund, he switched teams. I should have run a ‘unique’ filter before posting the list I guess!

Philosophil

G Money:
Next step: how we take these bands of players and use them to define competition. This is the current working definition (subject to almost-certain revision!):

A player is facing ‘top comp’ if at least one elite forward is on the ice, and no dreg forwards are on the ice

A player is facing ‘the dregs’ if at least one dreg forward is on the ice, and no elite forwards are on the ice

Any other combination of players is considered the ‘mushy middle’.

Now we slice and dice massive amounts of data to see how players do when facing these bands and voila – a next gen shot metric for comparing and contrasting players.

Here’s some initial test results on a handful of players.

Surprised that Yak is mushy middle. Dont have his stats at hand but iirc they are lower half on the team?

G Money

LadiesloveSmid,

It means he switched teams.

G Money

First thing to note here is you’ll see that a lot of players appear to be playing a lot against top comp. Part of that is because this is still a test version using a pretty lax definition of ‘top comp’. As that gets tightened down (WG and I intuitively think the list should be sixty players or less – average of about two per team), those numbers will shift away.

Some of it will inevitably remain, though, precisely because the best players tend to get the most time on ice, so they show up a lot (which is part of the effect that confounds the current QoC metrics, and which we hope to mitigate, though we’ll never get rid of it as it represents violence inherent in the system).

McD
Top Comp TOI 18339 CF 47.6 DFF 51.8
Muddle TOI 13593 CF 53.2 DFF 51.9
Dregs TOI 6090 61.2 DFF 69.7

McDavid faces a decent bit of the best competition already, though the gap you see here is actually quite modest. I’d call it ‘modestly sheltered’. Certainly, he rarely sees the easy competition. He’s mostly holding his own (47.6% CF and 51.8% DFF) even then – he’s a dangerous player no matter who is on the ice against him.

And he eats the dregs for breakfast, and spits out their bones.

Hall
Top Comp TOI 37878 CF 48.2 DFF 49.0
Muddle TOI 23785 CF 52.4 DFF 54.6
Dregs TOI 12713 CF 56.5 DFF 60.1

No surprises – Hall faces the best of the best night after night. And he holds his own (imagine what that number would look like with a consistently good D behind him). And like McDavid, he crushes the weak.

Sekera
Top Comp TOI 39771 CF 45.5 DFF 45.8
Muddle TOI 25021 CF 49.6 DFF 48.8
Dregs TOI 15739 CF 55.3 DFF 55.9

The TOI is as expected – like Hall, Sekera is fed to the wolves every night. Unlike Hall, he’s not holding his own, though. Perhaps a better partner, or easier competition.

Speaking of which – Sekera does just fine against the muddy middle, as expected. And manhandles the bottom of the roster.

Gryba
Top Comp TOI 20108 CF 47.4 DFF 48.0
Muddle TOI 16326 CF 49.8 DFF 45.5
Dregs TOI 12256 CF 52.7 DFF 50.5

This one does surprise me. Gryba faced more time against top comp (probably when he was with Sekera because of the righty shortage), and also does much better than expected against the ruffians of the world. The middle and dregs numbers make more sense to me.

A serviceable bottom pairing guy.

Nurse
Top Comp TOI 30354 CF 41.2 DFF 41.6
Muddle TOI 25345 CF 48.3 DFF 48.9
Dregs TOI 15341 CF 51.1 DFF 48.4

The Darnell Nurse story is writ large in these numbers. He’s spending too much time against top comp (the three numbers should be much closer together, like 26, 25, 20 instead of 30, 25, 15), and he’s getting eaten alive.

I guess a bit of good news is he’s muddling through against the muddle, and (mostly) holding his own on the bottom pairing. Next year is an important development year – a Top 4 D should be showing signs of being able to handle top comp by this point.

Korpikoski – I had to throw this one in. These numbers are about as sheltered as you can get, with Top and Middle neck and neck, with Dregs (the 10min TOI guys) modestly behind.

But it doesn’t matter who he faces, he generally gets crushed. And the WOWYs show he drags whoever he plays with into the mud with him.

Top Comp TOI 16990 CF 37.3 DFF 40.1
Muddle TOI 16510 CF 46.6 DFF 44.2
Dregs TOI 12661 CF 47.5 DFF 41.9

Addition by subtraction.

And last but certainly not least, Adam Larsson:

Top Comp TOI 46577 CF 42.3 DFF 44.5
Muddle TOI 24173 CF 45.8 DFF 53.2
Dregs TOI 14914 CF 48.9 DFF 51.8

Yes, he plays a crazy amount of top comp. Even moreso than Hall or Sekera, by quite a wide margin.

He does not do well – so I remain concerned that asking him to play top pairing here will require a pretty good developmental jump forward. We’d all be happier if he was slated at second pairing.

I am encouraged by his DFF numbers always being better than his CF numbers – the margin is wide enough that it doesn’t strike me as just random variation. Rather, to me it indicates that his reputation is earned, that he’s good at reducing the really dangerous stuff against.

And maybe with better forwards in front of him (except river pusher *sniff* Hall *sniff*), the ‘for’ side will go up and his numbers will show that much better.

I hope for the best but … I remain concerned.

LadiesloveSmid

frjohnk:
It’s Saturday night, I’ve had a few beers, sitting by the fire.

What the hell, just gonna throw some gas on the fire and say.

“Nail Yakupov will score more goals this year in the NHL than Taylor Hall will.
Both will play close to full seasons. ”

you have been on a roll tonight, I like to read it

LadiesloveSmid

G Money,

and is Granlund there 2x because he is so bad!? Calgary made off really well on that trade

LadiesloveSmid

Part of me hopes Davidson has a sophomore slump so he’s not on Las Vegas’s radar. What a wonderful player, and if they go out and get a scoring option on the back end there is a very small chance he is protected. Klefbom, Larsson, Davidson, Sekera is a top 4 I can live with and we can have Davidson dominating 3rd pair comp again if Fayne can break water against top 6 comp again with Sekera.

Someone asked me who my favourite player was recently and Davidson comes to mind. Plays the game so simple and so smart, fights for every inch he gets, and doesn’t take a second of ice time for granted. I recall WheatNOil posting something about Davidson’s successful breakouts or something of the ilk and it was something I had been admiring the entire latter half of the season. One time he went to what turned out to be the wrong side of the net because of how the puck was bouncing on a dump-in and smacked his stick against the ice, mad at himself for making the breakout a harder task. He still fended off the forward and advanced it, if I recall. Was a refreshing site from Jultz.

Water Fire

Looks good to me. Is Ladd there 2x because he’s so good?

You should go to sleep!

JimmyV1965

G Money:
1. A better QoC metric.

“Quality” in a player embodies many things.Yes, a good shot metric balance tells you a lot. But hardly everything.There’s the minor matter of scoring.Also how much trust the coach has can be reflected in TOI (but can also represent poor coaching or a shallow roster or both).

The first step, frankly, is to separate the measures we use for defenders and forwards. The scales of what they achieve are entirely different. Measuring them on the same scale is madness.

We’re starting with forwards.

The basic approach is to try and identify the elite forwards, the dreg forwards, and call everyone else ‘the mushy middle’.Those are the bands.

There are a number of approaches we have under consideration, but here’s the current test version (which isn’t perfect but is vastly better thanthe existing QoC metrics) for identifying elite forwards:

1. Start with forwards who’ve played at least one game over the last two years. n=641

2. Throw out any player without at least 41 games. n=475

3. Throw out any player with a CFRel < 0 (if they’re not in the plus on their own team, we assume they can’t be elite, no matter what). n=202

4. Keep only forwards who have an all situations (PP matters too) points/60 scoring rate that is in the upper third of all forwards from step #2.(That’s 1.92 p/60).n=110

5. And lastly, remove any forwards whose time on ice per game is less than the average time on ice for all forwards from step #2.If the coach distrusts that player to that extent, we should assume he’s not elite. n=95

So at this point we have 95 forwards, which is the top 15% of forwards from Step 1, and the top 20% of forwards from Step 2.So we can reasonably consider them in some sense ‘elite’.Do the results meet the sniff test?Here are the forwards (in noparticular order):

u’Pavel Datsyuk’, u’Daniel Sedin’, u’Nathan MacKinnon’, u’Ondrej Palat’, u’Nikolaj Ehlers’, u’Kyle Okposo’, u’Nikita Kucherov’, u’Marian Gaborik’, u’Claude Giroux’, u’Mike Ribeiro’, u’Gabriel Landeskog’, u’Mikael Backlund’, u’Evgeni Malkin’, u’Paul Stastny’, u’Jeff Carter’, u’Benoit Pouliot’, u’Jiri Hudler’, u’Max Pacioretty’, u’Connor McDavid’, u’Brendan Gallagher’, u’Jussi Jokinen’, u’Brad Marchand’, u’Kris Versteeg’, u’Vladimir Tarasenko’, u’Jamie Benn’, u’Blake Wheeler’, u’Aleksander Barkov’, u’Taylor Hall’, u’James Neal’, u’Mike Hoffman’, u’Zach Parise’, u’Vincent Trocheck’, u’Brandon Dubinsky’, u’Jakub Voracek’, u’Anze Kopitar’, u’Joe Pavelski’, u’Mark Stone’, u’Alex Ovechkin’, u’Patric Hornqvist’, u’Sidney Crosby’, u’Ryan Strome’, u’Loui Eriksson’, u’Jaden Schwartz’, u’John Tavares’, u’Ryan Getzlaf’, u’Jonathan Toews’, u’Tomas Plekanec’, u’Clarke MacArthur’, u’Rick Nash’, u’Bryan Little’, u’Filip Forsberg’, u’Patrick Sharp’, u’Jason Pominville’, u’Gustav Nyquist’, u’Sean Monahan’, u’Artemi Panarin’, u’Tyler Toffoli’, u’Mathieu Perreault’, u’Chris Kreider’, u’Jordan Eberle’, u’Leon Draisaitl’, u’Tomas Tatar’, u’Tyler Seguin’, u’Justin Williams’, u’Kris Versteeg’, u’Andrew Ladd’, u’Cam Atkinson’, u”Ryan O’Reilly”, u’Scott Hartnell’, u’Alex Galchenyuk’, u’Jaromir Jagr’, u’Nick Foligno’, u’Patrice Bergeron’, u’Mikko Koivu’, u’Kyle Turris’, u’Jiri Hudler’, u’Andrew Ladd’, u’Derick Brassard’, u’Mats Zuccarello’, u’Reilly Smith’, u’Nicklas Backstrom’, u’Johnny Gaudreau’, u’Marian Hossa’, u’Joe Thornton’, u’Henrik Sedin’, u’Radim Vrbata’, u’Mark Scheifele’, u’Jason Spezza’, u’Ryan Nugent-Hopkins’, u’Milan Lucic’, u’Brandon Saad’, u’Nino Niederreiter’, u’Martin Hanzal’, u’Frans Nielsen’, u’Wayne Simmonds’

Seem reasonable?Any obvious failures or omissions?

I’m sure there’s a few, but I can tell you, compare this list to the top of the traditional QoC list, and it is night and day.

Did you just blow my mind? Seriously though. Wow!!

G Money

Next step: how we take these bands of players and use them to define competition. This is the current working definition (subject to almost-certain revision!):

A player is facing ‘top comp’ if at least one elite forward is on the ice, and no dreg forwards are on the ice

A player is facing ‘the dregs’ if at least one dreg forward is on the ice, and no elite forwards are on the ice

Any other combination of players is considered the ‘mushy middle’.

Now we slice and dice massive amounts of data to see how players do when facing these bands and voila – a next gen shot metric for comparing and contrasting players.

Here’s some initial test results on a handful of players.

G Money

Now, to get the dregs of the league, reverse the steps above. Look for players in the bottom half of TOI, the bottom third in scoring, and (to get a bit more stringent) CFRel in the bottom third of all forwards.

In this case, any player who meets this criteria will be included, EVEN if they don’t have 41 games. After all the really shitty players might only get a cup of coffee, right?

That leaves us with 97 dreg forwards:

u’Brandon Prust’, u’Jordan Nolan’, u’Logan Shaw’, u’Drew Miller’, u’Andreas Martinsen’, u’Mike Richards’, u’Cedric Paquette’, u’Ben Smith’, u’Viktor Tikhonov’, u’Bobby Farnham’, u’Joe Vitale’, u’Jarret Stoll’, u’Byron Froese’, u’Matt Halischuk’, u’John Scott’, u’Ryan Reaves’, u’Max Talbot’, u’Tye McGinn’, u’Dennis Everberg’, u’Gabriel Bourque’, u’Iiro Pakarinen’, u’Luke Gazdic’, u’Chris Terry’, u’Anthony Peluso’, u’Miikka Salomaki’, u’Dainius Zubrus’, u’Craig Cunningham’, u’Eric Nystrom’, u’Daniel Paille’, u’Jared Boll’, u’Brandon Bollig’, u’Paul Gaustad’, u’David Legwand’, u’David Clarkson’, u’R.J. Umberger’, u’Quinton Howden’, u’Jacob De La Rose’, u’Austin Watson’, u’Boyd Gordon’, u’Chris Wagner’, u’Ben Smith’, u’Pierre-Edouard Bellemare’, u’Zach Sill’, u’Markus Granlund’, u’Joonas Kemppainen’, u’Chris Thorburn’, u’Viktor Tikhonov’, u’Travis Moen’, u’Jack Skille’, u’Landon Ferraro’, u’Tanner Glass’, u’Matt Hendricks’, u’Taylor Beck’, u’Derek MacKenzie’, u’Mike Brown’, u’Alex Chiasson’, u’Luke Glendening’, u’Dwight King’, u’Adam Cracknell’, u’Chris Wagner’, u’Nate Thompson’, u’Rene Bourque’, u’Brian Boyle’, u’Steve Ott’, u’Joakim Andersson’, u’Nicolas Deslauriers’, u’Curtis Lazar’, u’Michael Chaput’, u’Kyle Brodziak’, u’Mike Brown’, u’Chris VandeVelde’, u’Landon Ferraro’, u’Adam Cracknell’, u’Bobby Farnham’, u’Andy Andreoff’, u’Rob Klinkhammer’, u’Andrew Desjardins’, u’Dominic Moore’, u’Lauri Korpikoski’, u’Cody McLeod’, u’Brian Flynn’, u’Kevin Porter’, u’Michael Grabner’, u’Ryan White’, u’Brandon Mashinter’, u’Zac Rinaldo’, u’Barclay Goodrow’, u’Chris Porter’, u’Brad Malone’, u’Markus Granlund’, u’Jay McClement’, u’Ryan Carter’, u’Gregory Campbell’, u’Joakim Nordstrom’, u’John Scott’, u’Matt Stajan’

An interesting list to be sure. Surprised to see Stajan and Hendricks on there. Most of the others make sense.

JimmyV1965

AsiaOil:
Bruce McCurdy,

Bruce I doubt the fans who in some cases spent several hundreds of dollars to take their family to watch the $6 million dollar man thought the game meaningless, nor did the team that pays his salary. One of the biggest problems with this team has been the dreadfully low “give a shyte” readings of their best players on many nights. Winners don’t show up for work when they feel like it – they show up every damn day. Over the last 10 seasons, most of this team’s best nights came when a opponent for one reason or another didn’t show up and just them play river hockey. Serious teams in playoff mode simply obliterated them with almost zero pushback. Both TMac and Chia are determined to change that andmade progress last season. Guys who don’t buy in will be dealt – guys who do will get opportunities – we will see who falls into each group as the season progresses. If Ebs wants to stay he better learn to give 100% every night at both ends of the ice.

You’re right about that. However, the players would get just as discouraged as the fans. They are human too. And it must be hard knowing what an awful job mngt has done. Any coincidence Halls production dropped after they were out of the race? At least they made it until about Jsnuary this year.

G Money

1. A better QoC metric.

“Quality” in a player embodies many things. Yes, a good shot metric balance tells you a lot. But hardly everything. There’s the minor matter of scoring. Also how much trust the coach has can be reflected in TOI (but can also represent poor coaching or a shallow roster or both).

The first step, frankly, is to separate the measures we use for defenders and forwards. The scales of what they achieve are entirely different. Measuring them on the same scale is madness.

We’re starting with forwards.

The basic approach is to try and identify the elite forwards, the dreg forwards, and call everyone else ‘the mushy middle’. Those are the bands.

There are a number of approaches we have under consideration, but here’s the current test version (which isn’t perfect but is vastly better than the existing QoC metrics) for identifying elite forwards:

1. Start with forwards who’ve played at least one game over the last two years. n=641

2. Throw out any player without at least 41 games. n=475

3. Throw out any player with a CFRel < 0 (if they're not in the plus on their own team, we assume they can't be elite, no matter what). n=202

4. Keep only forwards who have an all situations (PP matters too) points/60 scoring rate that is in the upper third of all forwards from step #2. (That's 1.92 p/60). n=110

5. And lastly, remove any forwards whose time on ice per game is less than the average time on ice for all forwards from step #2. If the coach distrusts that player to that extent, we should assume he's not elite. n=95

So at this point we have 95 forwards, which is the top 15% of forwards from Step 1, and the top 20% of forwards from Step 2. So we can reasonably consider them in some sense 'elite'. Do the results meet the sniff test? Here are the forwards (in no particular order):

u'Pavel Datsyuk', u'Daniel Sedin', u'Nathan MacKinnon', u'Ondrej Palat', u'Nikolaj Ehlers', u'Kyle Okposo', u'Nikita Kucherov', u'Marian Gaborik', u'Claude Giroux', u'Mike Ribeiro', u'Gabriel Landeskog', u'Mikael Backlund', u'Evgeni Malkin', u'Paul Stastny', u'Jeff Carter', u'Benoit Pouliot', u'Jiri Hudler', u'Max Pacioretty', u'Connor McDavid', u'Brendan Gallagher', u'Jussi Jokinen', u'Brad Marchand', u'Kris Versteeg', u'Vladimir Tarasenko', u'Jamie Benn', u'Blake Wheeler', u'Aleksander Barkov', u'Taylor Hall', u'James Neal', u'Mike Hoffman', u'Zach Parise', u'Vincent Trocheck', u'Brandon Dubinsky', u'Jakub Voracek', u'Anze Kopitar', u'Joe Pavelski', u'Mark Stone', u'Alex Ovechkin', u'Patric Hornqvist', u'Sidney Crosby', u'Ryan Strome', u'Loui Eriksson', u'Jaden Schwartz', u'John Tavares', u'Ryan Getzlaf', u'Jonathan Toews', u'Tomas Plekanec', u'Clarke MacArthur', u'Rick Nash', u'Bryan Little', u'Filip Forsberg', u'Patrick Sharp', u'Jason Pominville', u'Gustav Nyquist', u'Sean Monahan', u'Artemi Panarin', u'Tyler Toffoli', u'Mathieu Perreault', u'Chris Kreider', u'Jordan Eberle', u'Leon Draisaitl', u'Tomas Tatar', u'Tyler Seguin', u'Justin Williams', u'Kris Versteeg', u'Andrew Ladd', u'Cam Atkinson', u"Ryan O'Reilly", u'Scott Hartnell', u'Alex Galchenyuk', u'Jaromir Jagr', u'Nick Foligno', u'Patrice Bergeron', u'Mikko Koivu', u'Kyle Turris', u'Jiri Hudler', u'Andrew Ladd', u'Derick Brassard', u'Mats Zuccarello', u'Reilly Smith', u'Nicklas Backstrom', u'Johnny Gaudreau', u'Marian Hossa', u'Joe Thornton', u'Henrik Sedin', u'Radim Vrbata', u'Mark Scheifele', u'Jason Spezza', u'Ryan Nugent-Hopkins', u'Milan Lucic', u'Brandon Saad', u'Nino Niederreiter', u'Martin Hanzal', u'Frans Nielsen', u'Wayne Simmonds'

Seem reasonable? Any obvious failures or omissions?

I'm sure there's a few, but I can tell you, compare this list to the top of the traditional QoC list, and it is night and day.

G Money

Howdy. About to post a bunch of stuff, most of which will make sense to WG, but I hope will be of interest to others.

Our current challenges:

1. Build a better blended metric to measure the ‘quality’ of players that goes into a QoC measure.

Currently this is CFRel or TOI, both of which on their own produce frankly terrible results.

The current ‘Corsiati’ pooh pooh the effect of QoC because the current QoC measures show minimal variance player to player. In other words, these folks claim, players overall just don’t face much difference in the competition they face, and QoC therefore doesn’t matter.

Do you believe them?

I don’t.

WG and I call 100% bullshit on that. In game coaching is almost entirely all about matchups – there’s lots of mythology in sport that has been disabused by analytics, but I run the numbers after every game and I can SEE the matchups happening.

So maybe just maybe it’s not QoC that’s irrelevant, it’s the metric we use to measure QoC that is bullshit, and we need better.

And frankly, it’s not hard to do. It’s a lot of work to code so that you can run it automatically on any player, but it’s not conceptually very hard.

2. Using this better metric, what’s the next step? The idea here (and this is heavily WG) is to separate competition into bands that pull out the players that are truly good and the truly bad, and then measure players (TOI, shot metrics) against those bands of players.

Then you can see who is *truly* facing the toughs or the dregs, and how often, and how they are doing.

This inherently produces a metric normalized for competition that can be compared apples to apples.

It is IMO a giant leap over one of the greatest flaws we face in using comparative shot metrics at the player level.

I’m this close *holds fingers a hair width apart* to finishing, but I have enough running with test output that I can’t sleep, so I figure I might as well get a head start on explaining some of this sh*t.

(There’s an additional method in the madness – as anyone involved in R&D will tell you, explaining what you’re doing helps not only to explain, but in your own head it helps to clarify the work done, the flaws, and where to go next)

So I am throwing the current development state intermediate results here for you to browse and abuse freely. (But I think they look pretty freaking good)

WG and I have chatted extensively on this stuff and he shies away from presenting this stuff at early stages because of the potential for misunderstanding, but

a. I’ll take it on faith that those who’ll actually read what I’m putting out here will by definition be a pretty smart crew and if you ask questions they’ll be more good than bad
b. as I said above, I find it helps the development process along.
c. WG is fast asleep and I’d say it’s a fair assumption that WG’s wife is probably gonna be pissed if I call him right now.

So here we go.

commonfan29

frjohnk: “Nail Yakupov will score more goals this year in the NHL than Taylor Hall will.
Both will play close to full seasons. ”

Yak will be healthy-scratched until he goes back to Russia.

This is because the team continues to be run by Old-School Hockey Guys, and OSHGs need to throw their weight around these days to show they’re still in charge.

commonfan29

hunter1909: Stelio Kontos:
Too many new people all at once. We need to build a wall before we end up like ON.
And KATZ is going to pay for it!

Katz doesn’t pay for things.

AsiaOil

Bruce McCurdy,

Bruce I doubt the fans who in some cases spent several hundreds of dollars to take their family to watch the $6 million dollar man thought the game meaningless, nor did the team that pays his salary. One of the biggest problems with this team has been the dreadfully low “give a shyte” readings of their best players on many nights. Winners don’t show up for work when they feel like it – they show up every damn day. Over the last 10 seasons, most of this team’s best nights came when a opponent for one reason or another didn’t show up and just them play river hockey. Serious teams in playoff mode simply obliterated them with almost zero pushback. Both TMac and Chia are determined to change that and made progress last season. Guys who don’t buy in will be dealt – guys who do will get opportunities – we will see who falls into each group as the season progresses. If Ebs wants to stay he better learn to give 100% every night at both ends of the ice.

Extend Russell

Omg wg I was just bugging man. No disrespect. I did read it. I read all your shit… Was bieng glib and silly.

Bruce McCurdy

Woodguy:
Hey Stephen Hawking,

Just read A Brief History of Time.

Like we needed a book to tell us that the universe is big and old.

*slow clap*

This made me laugh.

Didn’t Hawking also call out Eberle for missing a backcheck on a meaningless goal in a meaningless March game in another meaningless season? If he didn’t, he should have.

semi legendary rot lobster

ashley,

lobsters hv great friendship with yaks

lobsters also hv great cellys but humans never recognize this

therefore i have decided u r awesome i shall now call u ashlyobster

JimmyV1965

I know absolutely crap about crap, but I know the work of guys like Woodguy Is amazing. And I definitely feel good about Larsson thanks to the work he has done and I know we have a solid RHD to look forward to. Still got completely hosed in the deal though. Spit.

ashley

frjohnk:
It’s Saturday night, I’ve had a few beers, sitting by the fire.

What the hell, just gonna throw some gas on the fire and say.

“Nail Yakupov will score more goals this year in the NHL than Taylor Hall will.
Both will play close to full seasons. ”

No fire to gas over here. I would agree with this possibility.

For two years running, most of the comments on this blog have had Yakupov traded. Sometimes imminently…next 24 hours. I don’t know what it is about Yak, but something about him magnifies his flaws and hides his immense ability. He has shown us that ability and we should see those samples and see the value in him at a mere 2.5 million. The most striking was his play prior to the referee horse-tackle last year.

McDavid had lots of great things to say about playing with Yak (as did Roy the year before). He seems like a valuable piece. This is a domain where Hall has struggled…”playing off people” or playing to other’s style. Hall needs others to play to his style and RNH and Eberle had some success with that adjustment. McDavid struggled with a small sample.

It was interesting to hear Chia emphasize how well Lucic “plays off” other people when reporters were asking him what Lucic brings that Hall didn’t. Line chemistry is an important item which is what makes RNH so valuable. He looks good with everyone and makes everyone look good.

Yet there is something about Yak that people want to turf him. It seems personal or perhaps something about his personality. I can’t put my finger on it, but I was heartened to read all the positive comments on the thread from today and LT’s comments at the end of his post. Those comments are bang on. Trading Yak is ludicrous at this point.

Yak will have his day in the sun this year. Will we remember it next offseason?

From the biggest Yakcity fan. Yak4ever. I love that kid.

Raging inferno now? Is the house burning down?