Speculation: Roster Building Thread - Part XIII (Nanaki edition)

[MrAlmost]

Hoo boy

That was somehow less insightful then I thought it would be.

 

[lakeshirts37]

They might. He would make Ovechkin more effective on the powerplay with his goalie screens, which might make it worth it for the caps. That franchise’s number 1 objective for the next couple years is to get Ovy the goals record.

Lol they dont need Chris Kreider to get Ovi over the hump. And next couple of years?? He has 15 in 18 this year. He’ll break it this season even with missing time for injury or very early next season.

 

[Ruggs225]

please define "drive play"

this phrase gets tossed around constantly between analytics people and non analytics people and they often aren't talking about the same thing.

are you talking about contribution to xG%? ie possession quality? or are you just talking about elements of play and playstyle you notice when you watch him play?


For example: Machinehead does a thing where he uses different criteria to evaluate different players. For example, he posted Kreiders chart, adapted from evolving hockey numbers, which absolutely use possession quality metrics. Then he describes Kreiders effectiveness as a function of pure goal share - 58%. This is specifically an evaluation of results rather than process - entirely different from the chart.

I wonder what Zibanejad's possession quality and goal share metrics look like over the same span? Likely poor analytical process (xG%) but good results (GF%) - that's how he tends to profile. Zib, however, is a player that Machinehead currently castigates as a "problem player." It's true that Zibanejad's numbers, both XG% and GF%, have fallen this year - but as he himself said it's only 20 games. Clearly what differentiates 93 and 20 are their analytical profile rather than results - their xG%, etc. This analytical profile that describes quantity and quality of possession is what I use the umbrella term "play driving" to encompass. Kreider drives play, Zibanejad doesn't. Rangers own goal share when they share the ice... As Machinehead put it, Kreider carries Zib.

What's interesting, however, is that this standard does not apply universally. A great example is Kakko. Last year before the injury he played with 93/20 and their xg% was stellar. his, individually, too. The line just couldn't score, even though they didn't give up much either. It was almost like they were a perfect 3rd line being deployed against the opposing stars and fighting to a low-event draw.

Machinehead drew a few conclusions from this: 1 that instead of just being unlucky over a small sample size (as many of us asserted), Kakko was to blame: he actively prevented goals for because he's a dumb player. He entirely discredited the strong analytical profile in favor of the GF number. The metric for effectiveness changed entirely to suit the narrative he wanted to espouse, that Kakko sucked and was useless because he was allergic to scoring. (We are seeing now that this hypothesis was foolish).


So what's the point?

A. it matters to understand the terms someone else is using - especially one as both nebulous and essential to discourse as "play driving."

B. take everyone's opinion with a grain of salt - *especially* when it's presented with the type of smug self-assurance and refusal to self-evaluate that we see in this case

I am decidedly non analytics when it comes to hockey. I think most of the “advanced” stats are bullshit or doesnt really tell the full story. Baseball advanced stats though are very good, probably too good lol.

And its ironic that i literally do market research for a living and deal with surveys, reports, statistics daily.

When i talk about drive the play i mean the player who has the puck on his stick, the one who sees the ice and finds/creates a shooting lane or passing lane. Basically the person who is running the offense, which tends to more than likely a playmaker. Though certainly some people like Kakko drive the play by being great along the boards, and maintaining puck possession. So its a floating target, but the one constant is that a player must be good on the puck and have the puck alot to drive the play. If you arent good with the puck you arent driving the play.

Advanced stats can help tell a story, but too many use them to be the actual story which isnt true.

 

[Machinehead]

please define "drive play"

this phrase gets tossed around constantly between analytics people and non analytics people and they often aren't talking about the same thing.

are you talking about contribution to xG%? ie possession quality? or are you just talking about elements of play and playstyle you notice when you watch him play?


For example: Machinehead does a thing where he uses different criteria to evaluate different players. For example, he posted Kreiders chart, adapted from evolving hockey numbers, which absolutely use possession quality metrics. Then he describes Kreiders effectiveness as a function of pure goal share - 58%. This is specifically an evaluation of results rather than process - entirely different from the chart.

I wonder what Zibanejad's possession quality and goal share metrics look like over the same span? Likely poor analytical process (xG%) but good results (GF%) - that's how he tends to profile. Zib, however, is a player that Machinehead currently castigates as a "problem player." It's true that Zibanejad's numbers, both XG% and GF%, have fallen this year - but as he himself said it's only 20 games. Clearly what differentiates 93 and 20 are their analytical profile rather than results - their xG%, etc. This analytical profile that describes quantity and quality of possession is what I use the umbrella term "play driving" to encompass. Kreider drives play, Zibanejad doesn't. Rangers own goal share when they share the ice... As Machinehead put it, Kreider carries Zib.

What's interesting, however, is that this standard does not apply universally. A great example is Kakko. Last year before the injury he played with 93/20 and their xg% was stellar. his, individually, too. The line just couldn't score, even though they didn't give up much either. It was almost like they were a perfect 3rd line being deployed against the opposing stars and fighting to a low-event draw.

Machinehead drew a few conclusions from this: 1 that instead of just being unlucky over a small sample size (as many of us asserted), Kakko was to blame: he actively prevented goals for because he's a dumb player. He entirely discredited the strong analytical profile in favor of the GF number. The metric for effectiveness changed entirely to suit the narrative he wanted to espouse, that Kakko sucked and was useless because he was allergic to scoring. (We are seeing now that this hypothesis was foolish).


So what's the point?

A. it matters to understand the terms someone else is using - especially one as both nebulous and essential to discourse as "play driving."

B. take everyone's opinion with a grain of salt - *especially* when it's presented with the type of smug self-assurance and refusal to self-evaluate that we see in this case

Driving play means that your team is better when you're on the ice than when you're not. Any other use of it is incorrect.

What constitutes "better" evolves based on the context. The smaller the sample is, the more I lean towards expected numbers, possession numbers, and predictive indicators. After multiple seasons, the only thing that matters is goals, because why would anything else matter?

Kreider, over the course of his career, has been stellar at all of the above so the semantic arguments don't really apply to him. These 19 games have been an outlier for him in every measurable way, and frankly, it's not worth arguing over facts.

Also, the whole thing about "but Machinehead, Kakko is amazing at analytics so you should like him!" is a bit of a f***ing myth. This is his sixth season. In those seasons, his xGF% has been: good so far this year, actually terrible last year, excellent his second year, absolutely batshit f***ing AWFUL his rookie year, and decent the other two years.

It's been a mixed bag, and he does not have the resume of being north of 52% like clockwork that Kreider does. Analytics darling Kakko is something people mostly made up, which is not to say that he's bad at it, but he's not an Anthony Cirelli, with 40 points and a 55% xGF every year like people build him up to be.

 

[Ruggs225]

This team only has 3 holes.

Unfortunately those holes are all at the most important positions.

1C
1st pairing dman
2nd pairing dman

 

[Pawnee Rangers]

I am decidedly non analytics when it comes to hockey. I think most of the “advanced” stats are bullshit or doesnt really tell the full story. Baseball advanced stats though are very good, probably too good lol.

And its ironic that i literally do market research for a living and deal with surveys, reports, statistics daily.

When i talk about drive the play i mean the player who has the puck on his stick, the one who sees the ice and finds/creates a shooting lane or passing lane. Basically the person who is running the offense, which tends to more than likely a playmaker. Though certainly some people like Kakko drive the play by being great along the boards, and maintaining puck possession. So its a floating target, but the one constant is that a player must be good on the puck and have the puck alot to drive the play. If you arent good with the puck you arent driving the play.

Advanced stats can help tell a story, but too many use them to be the actual story which isnt true.

This was never Kreiders game though. Why expect that of him now.

 

[Machinehead]

This was never Kreiders game though. I why expect that of him now.

Yeah but that's also not what play-driving means, and I'm sorry if I wasn't clear on the term.

 

[IDvsEGO]

That was somehow less insightful then I thought it would be.

It was extremely hockey cliche.

I think he’s kinda calling Mika out here.

 

[Pawnee Rangers]

Yeah but that's also not what play-driving means, and I'm sorry if I wasn't clear on the term.

I agree with your definition

 

[Ruggs225]

Driving play means that your team is better when you're on the ice than when you're not. Any other use of it is incorrect.

What constitutes "better" evolves based on the context. The smaller the sample is, the more I lean towards expected numbers, possession numbers, and predictive indicators. After multiple seasons, the only thing that matters is goals, because why would anything else matter?

Kreider, over the course of his career, has been stellar at all of the above so the semantic arguments don't really apply to him. These 19 games have been an outlier for him in every measurable way, and frankly, it's not worth arguing over facts.

Also, the whole thing about "but Machinehead, Kakko is amazing at analytics so you should like him!" is a bit of a f***ing myth. This is his sixth season. In those seasons, his xGF% has been: good so far this year, actually terrible last year, excellent his second year, absolutely batshit f***ing AWFUL his rookie year, and decent the other two years.

It's been a mixed bag, and he does not have the resume of being north of 52% like clockwork that Kreider does. Analytics darling Kakko is something people mostly made up, which is not to say that he's bad at it, but he's not an Anthony Cirelli, with 40 points and a 55% xGF every year like people build him up to be.

Kakko has been +3, +9, +12, +6, and now is +13 the last five years.

So with him on the ice the team has consistently had more goals for than against. Therefore he has been a great play driver.

After all you said goals are the only thing that matters. And he is on for more goals for than against.

 

[DevilDog99]

we do.
That’s why we know that he’s the best defenseman we’ve had since leetch, and one of the few players on his team playing above his cap hit.

he’s a true 2 way defender, who generates offense and transitions the puck out of the zone defensively.
There are a few flaws to his game, as he can be overwhelmed by true elite forwards like mcdavid who not only move as fast as he thinks, but can also out muscle him physically.

He IS the best dman on our team right now. Leetch- he is not.

He could be better. The problem, as management and hopefully you guys, is attitude.

I am decidedly non analytics when it comes to hockey. I think most of the “advanced” stats are bullshit or doesnt really tell the full story. Baseball advanced stats though are very good, probably too good lol.

And its ironic that i literally do market research for a living and deal with surveys, reports, statistics daily.

When i talk about drive the play i mean the player who has the puck on his stick, the one who sees the ice and finds/creates a shooting lane or passing lane. Basically the person who is running the offense, which tends to more than likely a playmaker. Though certainly some people like Kakko drive the play by being great along the boards, and maintaining puck possession. So its a floating target, but the one constant is that a player must be good on the puck and have the puck alot to drive the play. If you arent good with the puck you arent driving the play.

Advanced stats can help tell a story, but too many use them to be the actual story which isnt true.

Thanks for this. Wish people would watch the games.

 

[Ruggs225]

This was never Kreiders game though. Why expect that of him now.

I know. Thats why i have consistently said he is not a play driver. I dont expect him of that now, which is also why i said he is completely replaceable and does more for this team in a trade that being on the ice.

Especially if we trade him for a top 4 dman.

 

[17futurecap]

Chris Kreider is the first player in NHL history to reveal what his injury is during the season.

 

[BleedBlueNYR]

Is Mollie Walker serious saying the Kreider article was “forthcoming”? He didn’t say anything of importance and then Mollie defends it by saying he gave details about his injury. She should be fired.

 

[Ruggs225]

Chris Kreider is the first player in NHL history to reveal what his injury is during the season.

Haha so true.

Wouldnt have guessed back spasms. But actually makes perfect sense.

 

[Pawnee Rangers]

I know. Thats why i have consistently said he is not a play driver. I dont expect him of that now, which is also why i said he is completely replaceable and does more for this team in a trade that being on the ice.

Especially if we trade him for a top 4 dman.

The guy holds or is close to the top of every franchise scoring record.

 

[Machinehead]

Kakko has been +3, +9, +12, +6, and now is +13 the last five years.

So with him on the ice the team has consistently had more goals for than against. Therefore he has been a great play driver.

After all you said goals are the only thing that matters. And he is on for more goals for than against.

Sure, that's a fair point.

 

[IDvsEGO]

Kakko has been +3, +9, +12, +6, and now is +13 the last five years.

So with him on the ice the team has consistently had more goals for than against. Therefore he has been a great play driver.

After all you said goals are the only thing that matters. And he is on for more goals for than against.

I’ve backed and defended Kakko for years now.

With that out of the way. STOP USING PLUS MINUS AS A MEASUREMENT!!

Jfc it’s a horrible measure of anything. It doesn’t take into account linemates, it doesn’t take into account anything other than goal/no goal.

Kakko is a puck possession monster, and because of that, he’s unlikely to be scored against because the other team uses up a ton of energy trying to get the puck away from him and his line.

Is Mollie Walker serious saying the Kreider article was “forthcoming”? He didn’t say anything of importance and then Mollie defends it by saying he gave details about his injury. She should be fired.

And this is the definition of clickbait. Which is why I usually avoid the post.
She did her job. She got fan engagement.
She might even get a raise for this one.

 

[Ruggs225]

I’ve backed and defended Kakko for years now.

With that out of the way. STOP USING PLUS MINUS AS A MEASUREMENT!!

Jfc it’s a horrible measure of anything. It doesn’t take into account linemates, it doesn’t take into account anything other than goal/no goal.

Kakko is a puck possession monster, and because of that, he’s unlikely to be scored against because the other team uses up a ton of energy trying to get the puck away from him and his line.

i was being facetious

The guy holds or is close to the top of every franchise scoring record.

That is more an indictment of how pathetic we have drafted forwards rather than a testament of how great he is

 

[Fitzy]

The NY post is about a half a standard deviation away from the national enquirer.

"See [celebrity] fashion mishap at the beach!"

No, since I'm not a 49 year old bored housewife, I don't think I will.

 

[Pawnee Rangers]

I’ve backed and defended Kakko for years

Jfc it’s a horrible measure of anything. It doesn’t take into account linemates, it doesn’t take into account anything other than goal/no goal.

Goals decide hockey games. That’s how we know who won or lost.

 

[will1066]

I’ve backed and defended Kakko for years now.

With that out of the way. STOP USING PLUS MINUS AS A MEASUREMENT!!

Jfc it’s a horrible measure of anything. It doesn’t take into account linemates, it doesn’t take into account anything other than goal/no goal.

Kakko is a puck possession monster, and because of that, he’s unlikely to be scored against because the other team uses up a ton of energy trying to get the puck away from him and his line.


And this is the definition of clickbait. Which is why I usually avoid the post.
She did her job. She got fan engagement.
She might even get a raise for this one.

Not specifically responding to you, but to me, it usually takes five guys to make a play other than having Connis McDavies or Nethan MacKinson skate through everyone and set up one of his boys.

Also, @Maliks PlusMinus in shambles now.

 

[IDvsEGO]

He IS the best dman on our team right now. Leetch- he is not.

He could be better. The problem, as management and hopefully you guys, is attitude.

Didn’t say he was better than leetch. Said best since.

And attitude is the most BS “I can’t articulate what I don’t like about this player” saying.

If you want to say he doesn’t throw hits, correct, that’s not his game.
If you want to say, he rarely beats a defender on a dump in, correct, he tends to just strip the puck after.

if you’re saying you watch the games, then I’d appreciate an actual hockey play that Fox doesn’t do well that’s leading to rangers losing?

 

[Pawnee Rangers]

The NY post is about a half a standard deviation away from the national enquirer.

"See [celebrity] fashion mishap at the beach!"

No, since I'm not a 49 year old bored housewife, I don't think I will.

Who doesn’t love a good celebrity nip slip?

 

[JimmyG89]

People give the roster a pass and assume it's just effort because this board has entirely failed to internalize how much scoring is up compared to the last time we contended.

Excellent point. Considering that scoring is up .3 goals per game per team. Each game has roughly a half goal more scored than from 2012-2017 seasons. You need a 50 point 3rd liner and two 40-45 point 3rd liners to be elite, with above average production in the top 6.

I don't really have too many issues with who is up front in this lineup, so long as they are living up to their end of the bargain. Too many right now are not. If those top 6 guys are, we're not talking like this.

The semi good news is that Kreider has been dealing with different kinds of back injuries all season. Anyone with a bad back knows it is compromising. I just wish the team and player wouldn't push it. Clearly they aren't with Chytil, but Kreider was being trotted out there.

Mika is unexplainable, similar to Trocheck.

Still, the biggest issue is puck movement from the back end. It's not just breakouts, but exits. Two guys are horrendous at exits AND don't just clear pucks out like others that struggle at it.

If Lindgren chipped off the glass or high flipped it to center ice, he'd be a much better player and at least a bottom pairing guy.