Roster Building Thread IV (2022-23): Luck of the Irish

[Glen Sathers Cigar]

The models themselves are also incredibly wrought with a subjective human opinion on how certain metrics should be weighted over others. And while I'd like to think those decisions when made initially were data-driven, they can still hold a ton of human error.

Long story short, if Patrick Kane garners a 3% WAR with us but is a PPG player on a Stanley Cup winning team, I'm not going to care one bit. Honestly, probably have the same reaction if he doesn't produce a single point. When push comes to shove, only one metric matters for a team - Stanley Cups.

I also say this as someone who had a stats emphasis in college and now works in Data Analytics, so this isn't an "old man yells at cloud" take. People simply overstate advanced analytics these days. They're way better than we used to have, but still nowhere close to what people think they are.

Bolded is very true and most of the legitimate stat guys wholly understand this, even if they may weight their models a little heavier because of their own biases, they understand that there's context and subjectivity and stuff.

The problem is people who don't fully understand the models and what they're doing just taking their chart screenshots and being like blue good/red bad and that's that.

I saw someone on twitter posting a WAR chart and making a definitive statement about some player and someone responded with some context around the player and why charts like that aren't always entirely accurate for this player or whatever and the person who posted the chart was like "Um it's literally just factual data you can't argue against it!" Like the person doesn't even understand that what they're looking at is not raw data it's a model's interpretation of the data - a model can be using accurate data but be a flawed model, that's entirely possible.

 

[mas0764]

Winning Cups is the only goal at the end of the day, but in the attempt to find out how to do that, if you think looking at the winning players have done or are doing is the best way to predict more winning, that is not a wise idea.

 

[McRanger92]

I think the more likely thing would be to send down Schneider and get someone around the same price. Lindgren being day to day instead of ending up on IR makes that more difficult. Oddly enough, if he had been hurt more and needed LTIR, we'd be able to add another.

The other issue with waiving Harpur is that some team that is cap strapped for next season might just claim him. I know for most we'd probably sign up for that, but at the same time, Gallant trust nobody else than the 7 guys they got right now.

If those two things happened, I believe they could add someone that currently has a 3.3M AAV, but you'd lose the ability to recall both before the postseason.

It would be a bad move to let the low possibility of losing Ben Harpur, who was on an AHL PTO 3 months ago, stop the team from adding another defenseman. I'm not going to die on this hill, but with the rate all these Eastern Conference teams are making moves, eliminating every variable could be the difference in the end. Everyone is so evenly matched but my 1 critique of the Rangers is that it will be tough for them to survive a major injury on defense.

 

[Kovalev27]

Wow. This really is their Last Dance with Bergeron & Krejci. Saturday is going to be a hell of a game.

I’ll be there too. Pumped

 

[fm]

These models can be informative but there are always gaps. I'm looking at a beloved model say that Mika's offensive output is 4th line level while his finishing is off the charts. How. How?

Meanwhile a guy like Kravtsov has good underling numbers so he's an analytics darling. But the model can't conceptualize the fact that he literally is too weak and frail to win a board battle or gain position in the o-zone.

Helpful tool but not an end-all be-all.

 

[Kovalev27]

Couldn’t u send Schneider down for a week and accumulate enough to get Braun on this roster at 50%? I mean u could demote Schneider tonight after the game to fit Braun immediately. Then we have what two games in the next 8 days? Wouldn’t we accumulate enough to bring him back up? We need another dman we really do. We need 8 minimum for the playoffs

 

[Raspewtin]

I've never in my life seen the vicious arms race that's going on in the East right now. Every team except Carolina is dramatically improving and paying an utter f***ton for it.

 

[McRanger92]

These models can be informative but there are always gaps. I'm looking at a beloved model say that Mika's offensive output is 4th line level while his finishing is off the charts. How. How?

Meanwhile a guy like Kravtsov has good underling numbers so he's an analytics darling. But the model can't conceptualize the fact that he literally is too weak and frail to win a board battle or gain position in the o-zone.

Helpful tool but not an end-all be-all.

I've never needed a statistical model to tell me if a team or player is really good or not. There's so much randomness and continuous play in the game of hockey compared to the other 3 major sports, I trust what I can actually see more.

 

[nsvoyageurs]

Everyone is trading their 2024 1st round picks, it must be going to be a crappy draft LOL

 

[Kovalev27]

Bruins gave up multiple picks including. A first for a guy with 4 goals. Jesus

 

[McRanger92]

I've never in my life seen the vicious arms race that's going on in the East right now. Every team except Carolina is dramatically improving and paying an utter f***ton for it.

Carolina is going to panic and trade for Kevin Hayes. Talk about a team that completely misread the market (intentionally or otherwise). They've had Patches' 7 mil burning a hole in their pocket for 2 months and all they ended up with was another analytics darling who can't score and an offensive defenseman who has never contributed meaningfully to a contender. Rooting for the Devils the rest of the way. We get a Carolina team we match up well with in the first round, and the youthful Devils get a battle tested Islanders or Penguins squad.

 

[JimmyG89]

It would be a bad move to let the low possibility of losing Ben Harpur, who was on an AHL PTO 3 months ago, stop the team from adding another defenseman. I'm not going to die on this hill, but with the rate all these Eastern Conference teams are making moves, eliminating every variable could be the difference in the end. Everyone is so evenly matched but my 1 critique of the Rangers is that it will be tough for them to survive a major injury on defense.

Neither would I, but I'm trying to piece this together how Drury would be looking at it.

If we could just waive Harpur today, he can still play tonight and they can still emergency re-call Hajek/Robertson, you're opening up an option like Kulikov, but you're running 6D the rest of the way, can't emergency re-call Schneider.

I'm with you that at this rate, you go out and get one more defender on the cheap and make sure for G1 that you have Harpur being the 8th option.

 

[NYR Viper]

Boston continues to load up on guys who are a PIA to play against...

 

[Kovalev27]

You know whose deal is going to look like a steal like yesterday even. Adam Fox.

Guys are going to start making 11 plus like pasta just got.

Even Mika’s deal already looks like a steal when u see Horvat and Larkin making 8.7 mil

 

[EpicDing]

 

[McRanger92]

It's a bit of a relief knowing we added 2 stars who have vanquished the Bruins in the Stanley Cup. They've had our number for as long as I can remember but I think it could be different this year if we do face them

 

[NYR Viper]

You know whose deal is going to look like a steal like yesterday even. Adam Fox.

Guys are going to start making 11 plus like pasta just got.

Even Mika’s deal already looks like a steal when u see Horvat and Larkin making 8.7 mil

The same thing will happen to the mid-tier guys. Even guys like Goodrow will have team friendly deals shortly.

 

[Kovalev27]

Bruins have a couple injuries that may be an issue hall and foligno. Wonder if this has much to do with that.

 

[Avery16]

For all the complaints about the cap, we and other teams were able to pull off moves that would seem unlikely even back in the 90s. Just comes down to smart cap management. Can make or break an org. We've had a taste of both in recent years.

 

[duhmetreE]

Bruins are absolutely stacked.

We're right below them right now but our offense has much more potential

Ullmark might break in the playoffs. Hopefully they get the Islanders or Pittsburgh.

 

[JimmyG89]

Couldn’t u send Schneider down for a week and accumulate enough to get Braun on this roster at 50%? I mean u could demote Schneider tonight after the game to fit Braun immediately. Then we have what two games in the next 8 days? Wouldn’t we accumulate enough to bring him back up? We need another dman we really do. We need 8 minimum for the playoffs

That was a question that I did have, I figure you can still accrue cap space after the deadline if you are under it. Schneider down after tonight's game, trade for a D with less AAV than his ELC, and then call him up once you can fit him in, but I am not sure how long that would take.

 

[DanielBrassard]

If they are healthy Bertuzzi is on a 3rd line with Hall and coyle. That is just a ridiculous amount of depth.

 

[Avery16]

Braun isn't any good. He wasn't that great for us. I'd rather have a week of Schneider than Braun for the entire playoffs. Is it worth it for another option? Maybe.

 

[DanielBrassard]

That was a question that I did have, I figure you can still accrue cap space after the deadline if you are under it. Schneider down after tonight's game, trade for a D with less AAV than his ELC, and then call him up once you can fit him in, but I am not sure how long that would take.

You do accrue cap space until the end of the season. But I would do this as well, let's be honest it's not like Schneider is playing all that well. Someone like TVR would be an upgrade if you fit him in.

 

[MrAlmost]

The guy whose deal looks the best as we keep going forward is Trocheck.