Roster Building/Team Building/Future Trade/Drafting thread.

[Kurtz]

You brought in Marchand, Pageau, Colton, Jeannot, Tanev, Brown, Kreider at the expense of Marner, Tavares, Kampf, McMann, Holmberg, Robertson, Lorentz.

Sorted by EV D impact last season:

View attachment 1047772

You get marginally worse. Primarily lead by a much worse 4th line defensively.

Are you ok with an output that for example tells you that Robertson is a better 5 on 5 defensive player than Marner?

 

[thusk]

So your choice is Laughton for 2nd. line center?

At least he is real.

If you sign a winger who fit with the team much more than any C available? yeah can be a possibility

Honestly for me a Marchand-Laughton-Nylander or Kreider laughton-Nylander would be better than a duchene-granlund-Nylander line and will cost less

It's nothing against duchene or Granlund, they are pretty good player... i just don't any fit with their game playing alongside Nylander in the Berube system.

 

[Gaberd2608]

If you sign a winger who fit with the team much more than any C available? yeah can be a possibility

Honestly for me a Marchand-Laughton-Nylander or Kreider laughton-Nylander would be better than a duchene-granlund-Nylander line and will cost less

It's nothing against duchene or Granlund, they are pretty good player... i just don't any fit with their game playing alongside Nylander in the Berube system.

I don't think Laughton can keep up with Nylander. Most people like him as a 4c not a 2c. He lacks the offensive creativity and skill to play effectively with Nylander imo. Marchand I think has lost a step speed wise to be effective with Nylander. Domi may be a better fit with Marchand and Laughton on the 3rd line as a pest line.

I actually think Granlund would be ok with Nylander. Slot in another fast skilled winger and maybe ok

 

[Americanadian]

Are you ok with an output that for example tells you that Robertson is a better 5 on 5 defensive player than Marner?

It's telling me Robertson had better defensive results than Marner (5v5 GA/60 and xGA/60), especially relative to his line mates, which he did.

Numbers don't measure who has the best ability, they measure what happened.

 

[Fogelhund]

I don't think Laughton can keep up with Nylander. Most people like him as a 4c not a 2c. He lacks the offensive creativity and skill to play effectively with Nylander imo. Marchand I think has lost a step speed wise to be effective with Nylander. Domi may be a better fit with Marchand and Laughton on the 3rd line as a pest line.

I actually think Granlund would be ok with Nylander. Slot in another fast skilled winger and maybe ok

Ehlers would make for a great, speedy and dangerous line... It might also make it unaffordable.

 

[Dreakmur]

You’d be hard pressed to find a data centered person who thinks Kyle Dubas did a good job.

Can’t enter a thread on this board without tripping over those guys.

 

[Dreakmur]

It's telling me Robertson had better defensive results than Marner (5v5 GA/60 and xGA/60), especially relative to his line mates, which he did.

Numbers don't measure who has the best ability, they measure what happened.

Measuring what happened would be great, but it doesn’t do that either. It only measures a handful of data points that give at best a partial picture.

 

[Americanadian]

Can’t enter a thread on this board without tripping over those guys.

They aren’t data guys.

 

[Kurtz]

It's telling me Robertson had better defensive results than Marner (5v5 GA/60 and xGA/60), especially relative to his line mates, which he did.

Numbers don't measure who has the best ability, they measure what happened.

You're attempting to use them for their predictive value, right? You are leaning on a composite stat that tells you that Robertson was better 5 on 5 defensively than Mitch (note you didn't present any alternative context - you just filtered on that stat). Knowing this, do you feel that this stat is worth leaning on in making future assesment of player defensive value?

Would it not be wiser, at the very least, to filter for multiple seasons to smooth out such obvious outliers? Or are you comfortable with the assertion that Mich is a below-average 5 on 5 defensive player... because that's a big claim.

 

[Americanadian]

Measuring what happened would be great, but it doesn’t do that either. It only measures a handful of data points that give at best a partial picture.

The data that is used to create these “advanced” numbers are just a sum out goal and shot outcomes.

It doesn’t measure if a player is in the right position or how many stick lifts they did. It measure how many goals went in the net when they were on the ice throughout the entire season and where those shots were taken.

 

[Americanadian]

You're attempting to use them for their predictive value, right? You are leaning on a composite stat that tells you that Robertson was better 5 on 5 defensively than Mitch (note you didn't present any alternative context - you just filtered on that stat). Knowing this, do you feel that this stat is worth leaning on in making future assesment of player defensive value?

I’m using this number to say that the proposed players going in were not better defensively last year than the proposed players going out.

Would it not be wiser, at the very least, to filter for multiple seasons to smooth out such obvious outliers? Or are you comfortable with the assertion that Mich is a below-average 5 on 5 defensive player... because that's a big claim.

For a Marner vs Robertson debate absolutely (and it’s not a debate).

When we’re talking about Marchand and Kreider and Pageau, all guys in their mid 30s, you need to look at the latest data because they’re on the decline.

 

[thusk]

I don't think Laughton can keep up with Nylander. Most people like him as a 4c not a 2c. He lacks the offensive creativity and skill to play effectively with Nylander imo. Marchand I think has lost a step speed wise to be effective with Nylander. Domi may be a better fit with Marchand and Laughton on the 3rd line as a pest line.

I actually think Granlund would be ok with Nylander. Slot in another fast skilled winger and maybe ok

Nylander don't need player to make play, he need player who gonna win puck battle and give him the puck. Nylander will hurt you if he have the puck on his stick

I will make a comparable with Bunting. He don't have a lot of offensive skill but working well with Matthews/Marner because he was working extremely hard, winning battle and bringing the puck back to matthews or marner. Alone bunting bring basically nothing offensively, when knies step in his spot on top line he became useless for leafs... but if he become the 3rd wheel of a line, yeah he can help and be surprising. The best player are not always the guy with the modt offensive skills

 

[Kurtz]

I’m using this number to say that the proposed players going in were not better defensively last year than the proposed players going out.

For a Marner vs Robertson debate absolutely (and it’s not a debate).

When we’re talking about Marchand and Kreider and Pageau, all guys in their mid 30s, you need to look at the latest data because they’re on the decline.

Gotcha. Fair enough on age-related decline, but I'd be very hesitant to lean on any stat that presents Marner as a below-average defensive player 5 on 5.

 

[Legion34]

B. Tkachuk getting 8.2Mx7 (9.6M today) after putting up a career high 53 point pace in his 3rd season.

Josh Anderson getting 5.5Mx7 (6.4M today) with 1 27G/47P season under his belt, despite only hitting 29 points 2 other times.

Tom Wilson getting 5.17Mx6 (6.2M today) with a career high of 14G/35P.

I don't know how many other real stylistic comparables there are for Knies.

Eichel got 10 x8 he didn’t ll

I dont see Bert getting bought out, but if so sign me up. I think you'd get a lot more value and milage out of a lineup that has Matthews playing with 2 cheaper guys like Bertuzzi and Domi, a Knies-Nylander 2nd line that has the center being the worst player on that line, a 3rd line that has Cowan learning from Tavares, and then a checking line featuring Tanev-Laughton.

Bertuzzi-Matthews-Domi
Knies-X-Nylander
Cowan-Tavares-Pacioretty
Tanev-Laughton-McMann

If he was going to be bought out. Wouldn’t we just take him for free?
Doesn’t Chicago need to spend tons just to get to the floor?

 

[ITM]

Everyone had the winner of Toronto vs Florida going to the Cup finals, and if Stolarz wasnt 'injured' it would have been us.

I despise that everyone downplays this team to be less than it is, because we play in the most stacked division in the NHL.
whoever survives the Atlantic typically is in the Finals.

I’m not exactly sure what everyone thinks. But I think it’s public knowledge that it’s unlikely Bennett and Marchand join the team they beat, by leaving the team that beat them.

I suppose there’s a Lloyd Christmas chance…?

 

[58 Years No Cup]

Measuring what happened would be great, but it doesn’t do that either. It only measures a handful of data points that give at best a partial picture.

He's apparently unaware of the GIGO Principle.

 

[Kurtz]

He's apparently unaware of the GIGO Principle.

The data itself may not be garbage, but the weights they are applying to the formula variables might be off (judging solely by that limited output of a dozen or so guys in a single season). Maybe if we were to filter for multiple seasons and a bigger player sample we can get a more reliable output.

 

[Americanadian]

The data itself may not be garbage, but the weights they are applying to the formula variables might be off (judging solely by that limited output of a dozen or so guys in a single season). Maybe if we were to filter for multiple seasons and a bigger player sample we can get a more reliable output.

When I am projecting I use 3 years weighted at 50/30/20. When I’m looking at what happened last year I use the data from last year.

 

[Kurtz]

When I am projecting I use 3 years weighted at 50/30/20. When I’m looking at what happened last year I use the data from last year.

I don't mean recency-based weights, I mean the weights of the individual variables that go into that composite EVD stat.

For example, do you see how your chart seems to love bottom-six guys but not the top-6 guys? That tells me that the weight of the QOC variable (which I believe is one of the variables being used in the stat) is too light.

 

[Menzinger]

Can’t see a world in which Bennett or Marchand discard winning ways and sound judgement for Toronto, given their play and in particular, Marchand’s comments.

I can see Zito finessing Florida’s structure to offer Marchand a retirement contract.

I think both are going to prioritize money at this point. This is Bennets chance to set himself up for the rest of his life and Marchand is looking for a nice retirement contract. I'm not sure Florida will have the room for both, especially if Marchand is looking for 3-4 years at 8ish mil

Florida either pays up or Tre will have his chance to make a pitch. Now does it make sense for the Leafs to offer those kinds of deals becomes another question all together

 

[Americanadian]

I don't mean recency-based weights, I mean the weights of the individual variables that go into that composite EVD stat.

For example, do you see how your chart seems to love bottom-six guys but not the top-6 guys? That tells me that the weight of the QOC variable (which I believe is one of the variables being used in the stat) is too light.

It’s not my data, I’m using EvolvingHockey. If you were to sort out the whole league it favors top end players. The top 6 players in question (Marchand, Kreider, Marner) just didn’t grade out well defensively this year.

 

[horner]

RFA's
Which ones aren't going to be offered a contract and teams will walk away.

2026 & 2027
Is there a Holloway or Broberg out there it cost St Louis a 2nd and a 3rd

 

[Dreakmur]

The data that is used to create these “advanced” numbers are just a sum out goal and shot outcomes.

It doesn’t measure if a player is in the right position or how many stick lifts they did. It measure how many goals went in the net when they were on the ice throughout the entire season and where those shots were taken.

That’s exactly what I said - handful of data points that shows a partial picture.

 

[Dreakmur]

RFA's
Which ones aren't going to be offered a contract and teams will walk away.

2026 & 2027
Is there a Holloway or Broberg out there it cost St Louis a 2nd and a 3rd

With a rising cap, it’s pretty unlikely.

 

[Americanadian]

That’s exactly what I said - handful of data points that shows a partial picture.

It measures the outcome. When your sample is a whole season you can make conclusions on the outcome.