Devils team discussion (news, notes and speculation) - 2022-23 season thread part III

[Guttersniped]

 

[Guttersniped]

 

[Better Call Sal]

am i reading that right? 6.8M for zacha? get it done, sweeney!

Do you not have to eat a bag of shit if he tops 50 points in Boston?

 

[Guttersniped]

 

[MartyOwns]

Do you not have to eat a bag of shit if he tops 50 points in Boston?

no lol

 

[Clam Jensen]

View attachment 620481
View attachment 620482
View attachment 620480View attachment 620483View attachment 620484View attachment 620485

This is eye porn and a lot stands out, but wow Fabian is really something else.

 

[Camille the Eel]

You’re looking at the “market value” and not the whole card and yes, it’s not great at measuring defensive stick work. The rating is affected by his two previous seasons too.

Marino is definitely overrated on this board for being the anti-Severson. I overrate him for that but he’s also what we need and he’s only played 27 games here.

People are getting touchy but Marino plays brutal minutes and his numbers suffer for it. The rating change over the year too.

Also I doubt that the rating captures the truly game saving quality of some of the plays. The last second break up of a sure tap in of a cross ice pass at the back post, that we saw; the coming out of nowhere to run down Kadri and prevent a break away . . . The importance of some of his plays and timing in the game, I can’t see how the models easily capture that. I’d be surprised if they do.

 

[Jets012]

Marino is a nice player. I think the model is more than fair to him. The issue is he got SO much praise to start the year that it caused him to be a bit overrated which is why his number looks so low in The Athletic model. Maybe it's because he has a handful of good defensive plays that are memorable and he was the new guy who knows.

Anyone who thought he was on a similar level to Siegenthaler were vastly mistaken though. That's more of a testament to how go Siegs is though. An absolute stud in every definition.

 

[SteveCangialosi123]

The model overrates players getting cake minutes and doing well in them and doesn’t seem to know how to rate a guy getting very difficult assignments and doing okay in them.

I’m sure this model would have Butcher-Lovejoy as some force when they were together his rookie year.

 

[Stephen Gionta]

The "market value" projections in those models mean absolutely nothing in the real world.

Any model that says Fabian Zetterlund's market value is 7 million should be immediately dismissed.

The model overrates players getting cake minutes and doing well in them and doesn’t seem to know how to rate a guy getting very difficult assignments and doing okay in them.

I’m sure this model would have Butcher-Lovejoy as some force when they were together his rookie year.

Yup.

That model has zero context for anything at all. Again, why it is complete nonsense.

This is eye porn and a lot stands out, but wow Fabian is really something else.

You really shouldn't be taking these models seriously like at all. They mean absolutely nothing and are essentially just made up.

They take one aspect of a players performance and put a "value" and a contract number on it. It's nonsense.

 

[Saugus]

Those market valuations seem incredibly out of whack for a world where the salary cap exists. Even if Jack Hughes hit the open market tomorrow as a UFA, nobody would ever pay him the $15 M that model claims he is worth.

As a relative measure for comparing players against each other with regards to the value they generate for their teams, it might be fine. But it should not be taken as an objective metric for a player's absolute worth, because it is clearly not giving realistic results.

 

[Guttersniped]

Also I doubt that the rating captures the truly game saving quality of some of the plays. The last second break up of a sure tap in of a cross ice pass at the back post, that we saw; the coming out of nowhere to run down Kadri and prevent a break away . . . The importance of some of his plays and timing in the game, I can’t see how the models easily capture that. I’d be surprised if they do.

It really liked Marino defensively last season this year.

The value of his hard minutes are apparently blunted by us being good now and our schedule being soft so far.

Add to his lack of offensive value and you get this.

These results can change over time too. If you’re saying this doesn’t quantify defensive value perfectly, then sure. That’s hard to do. It’s also hard to do just with the eye test.

 

[Triumph]

Marino gets the hard minutes but he has been underwater since the game against Calgary at home. The Devils shouldn't be being beaten in tough minutes unless Marino is also playing a lot with McLeod. Before that game, Marino was up with the rest of the D.

 

[Camille the Eel]

It really liked Marino defensively last season this year.

The value of his hard minutes are apparently blunted by us being good now and our schedule being soft so far.

Add to his lack of offensive value and you get this.

These results can change over time too. If you’re saying this doesn’t quantify defensive value perfectly, then sure. That’s hard to do. It’s also hard to do just with the eye test.

I’m not against models and analytics. They’re very useful tools and understanding their limitations and biases makes them even more useful. It’s really interesting to go into how the analysis was modeled - based on what data and how weighted and constructed. Every once in a while when you see a truly anomalous and anti intuitive result going into that stuff can explain it.

 

[Guttersniped]

Those market valuations seem incredibly out of whack for a world where the salary cap exists. Even if Jack Hughes hit the open market tomorrow as a UFA, nobody would ever pay him the $15 M that model claims he is worth.

As a relative measure for comparing players against each other with regards to the value they generate for their teams, it might be fine. But it should not be taken as an objective metric for a player's absolute worth, because it is clearly not giving realistic results.

Market Value here is

Based on contract projections for every player in the league provided by Evolving Hockey compared to projected future value measured using GSVA, I estimate that the value of a win on the open market for a forward is $3.27 million*, with replacement level being worth just above $1.1 million. For defense, it’s $3.5 million and $1.75 million respectively.

When he says Hughes is worth 15m or McDavid is worth 21.2m he’s not predicting or suggesting their real-life salary.

It’s just how value functions in a cap system. Top stars are worth significantly more than lower tier players but their salaries can’t be that disproportionate.

It’s even more apparent in the NBA where the top stars who carry the teams are comparatively underpaid in relation to the lower tier complimentary (or worse) players who make a ton while contributing significantly less.

We’re winning so we have a lot guys with value, it’s a very results oriented model.

The price of a win and why it means Connor McDavid won’t always be the NHL’s highest-paid player

Whoever passes McDavid will be worthy of the money, another sign that McDavid is very underpaid.

theathletic.com

 

[MadDevil]

What I take from all those is we're getting pretty good value out of everybody except Blackwood, which seems about right.

 

[Eggtimer]

All our player cards show everyone but Blackwood and McLeod as beig bargain contracts relative to their value on ice .
Only one card I don’t get / think they under value is Marino. Guy has been a stud for us, I was expecting way higher ratings

 

[MadDevil]

Rotfl at this sheer idiocy.

A ten year old mega fan with a pencil and notebook.

Hey guys, just bc someone comes up with a formula to crunch numbers and writes a little code doesn't immediately give it credence. The fact that The Athletic lends this thing any sort of hint of veracity says more about their stupid takes than it does about our team.

Is there any stat or visual representation of data or model that doesn't trigger you?

 

[Camille the Eel]

I also think the NHL salaries are way out of line with other major sports. If hockey is really a major sport in the US. The networks (broadcasts) don’t pay as much for the sport and it all apparently works back from there into cap figures as a %? Something like that anyway is my hazy understanding of how it’s calculated.

 

[Camille the Eel]

Is there any stat or visual representation of data or model that doesn't trigger you?

The Athletic is The NY Times and like all web content there’s a premium on just getting you to click. It needs to be intriguing and click bait, it doesn’t need to be particularly well reasoned or enlightening. It’s intellectual entertainment. There are a bunch of posters here to whom I would pay more credence and who are more interesting than anything in that web mag.

The Times has really gone down hill.

 

[Stephen Gionta]

Those market valuations seem incredibly out of whack for a world where the salary cap exists. Even if Jack Hughes hit the open market tomorrow as a UFA, nobody would ever pay him the $15 M that model claims he is worth.

As a relative measure for comparing players against each other with regards to the value they generate for their teams, it might be fine. But it should not be taken as an objective metric for a player's absolute worth, because it is clearly not giving realistic results.

Yep.

If you take all 32 teams, and multiply that by 81.5m - that is the total amount of salary cap space available in the league. Which comes to $2,608,000,000.

I would be willing to bet if you took every one of these model contract projections and added them together, they would VASTLY eclipse that number. Which goes to show how unrealistic and outlandish these numbers are.

Is there any stat or visual representation of data or model that doesn't trigger you?

He's right though..

 

[SteveCangialosi123]

I do not think these models can accurately determine who is driving results either. Like if Marino was with a left handed Dougie Hamilton and Siegenthaler was with a right handed Ryan Graves, how do we think those numbers would be looking for both of those guys?

 

[TrufleShufle]

I do not think these models can accurately determine who is driving results either. Like if Marino was with a left handed Dougie Hamilton and Siegenthaler was with a right handed Ryan Graves, how do we think those numbers would be looking for both of those guys?

Very different, we would have like 8 D man (exluding Bahl).... unless you are saying we cut left handed Graves and right handed Hamilton. Or they could play together and cut Sevs and Smith. But what would you do for jersey plates? R. Hamilton and L. Hamilton?

 

[SteveCangialosi123]

Very different, we would have like 8 D man (exluding Bahl).... unless you are saying we cut left handed Graves and right handed Hamilton. Or they could play together and cut Sevs and Smith. But what would you do for jersey plates? R. Hamilton and L. Hamilton?

Reminds me of this scene

 

[ZachaFlockaFlame]

I mean Marino has been bad the last few games and we're still early on where a few bad games can tank a player card.