Prospect Info: Devils picking 4th Overall.

[Unknown Caller]

7% Star probabilty lmao, give me a break. Owen Power is, and will be a stud. Use your eyes and watch the games. will he be a top 5 defensmen in the NHL? I dont know, maybe not. Drafting is about projection and hes just as close to a projected #1 defensman as any Ive seen in the last 5 years, so if im Buffalo im grabbing him and not thinking twice.

This is just not accurate. Even a traditional eye test guy like Chris Peters said this week that he would have Power around #10 in 2019 (behind Byram and Sieder and around Broberg) and #7 last year (behind Drysdale and Sanderson). How Power is now ahead of not only them, but guys like Dahlin and Makar and Heiskenen, is beyond me.

 

[BurntToast]

Can I vote for all options. Seriously, I am happy that I’m not Fitz.

@StevenToddIves

I just listened to the new “Speak of the Devils” podcast with Morreale and Kournianos;

Morreale hit me with some insight, Luke Hughes plays both sides? Plus, he felt Luke was better defensively, while Clark is more offensive. (At this point, I would lean Hughes)

I have been doing my best not getting attached. Power would be a lock with any other team than Buffalo (That organization is in Chaos). The next 9-10 picks are anyones guess.

 

[Blackjack]

Can I vote for all options. Seriously, I am happy that I’m not Fitz.

@StevenToddIves

I just listened to the new “Speak of the Devils” podcast with Morreale and Kournianos;

Morreale hit me with some insight, Luke Hughes plays both sides? Plus, he felt Luke was better defensively, while Clark is more offensive. (At this point, I would lean Hughes)

I have been doing my best not getting attached. Power would be a lock with any other team than Buffalo (That organization is in Chaos). The next 9-10 picks are anyones guess.

I wouldn't read too much into that. Smith played RD for Spokane because they needed RD, I have every expectation that Hughes will be a full time LD in the NHL.

 

[devilsblood]

This is just not accurate. Even a traditional eye test guy like Chris Peters said this week that he would have Power around #10 in 2019 (behind Byram and Sieder and around Broberg) and #7 last year (behind Drysdale and Sanderson). How Power is now ahead of not only them, but guys like Dahlin and Makar and Heiskenen, is beyond me.

Of your list, Dahlin was the one guy who looked like a sure thing #1 d-man pre draft. Guys like Makar and Heiskanen, have considerably raised their prospects post draft, but I don't think there was a pre draft consensus that they would be #1's.

Peters is one opinion, and it should be considered, but Power is pretty consensus the top player in this draft, and a #1 pick is very much expected to be a #1 d-man.

 

[billingtons ghost]

If Power has success, his star % will increase, if he does not, it will decrease.

I can't believe you used 'star %' and 'ludicrous' in the same post not pertaining to each other.

Bader's SnakeOil% is 939% according to my formula. I saw his Framus score as low as 13 Grommets in some sites.

It's hard to believe someone could:
1. Quantify and distill the proper formula for someone being a 'star'.
2. Have sufficient data on these young players to make any meaningful prediction.
3. Do this during COVID, with even less data.

My biggest problem is slapping a percent sign behind an opinion just so others will swallow the hook.

 

[Triumph]

I can't believe you used 'star %' and 'ludicrous' in the same post not pertaining to each other.

Bader's SnakeOil% is 939% according to my formula. I saw his Framus score as low as 13 Grommets in some sites.

It's hard to believe someone could:
1. Quantify and distill the proper formula for someone being a 'star'.
2. Have sufficient data on these young players to make any meaningful prediction.
3. Do this during COVID, with even less data.

My biggest problem is slapping a percent sign behind an opinion just so others will swallow the hook.

It doesn't matter what you or I think, the question is if he beats the market. If his predictions, ordered every year, beat the NHL draft, then he is on to something even if that tool is not particularly helpful at picking an individual player. As I've said upthread, I would guess this beats the NHL draft in aggregate, but only very slightly, and again, that when something like this misses, they tend to miss bigger than NHL teams miss these days. It's not very useful to NHL teams except as a way of comparing players' numbers across leagues which can be quite a challenge given that there's ~10+ leagues represented in the first 3 rounds of the NHL draft. It can help you look at the players your team should want to draft.

 

[Unknown Caller]

Of your list, Dahlin was the one guy who looked like a sure thing #1 d-man pre draft. Guys like Makar and Heiskanen, have considerably raised their prospects post draft, but I don't think there was a pre draft consensus that they would be #1's.

Peters is one opinion, and it should be considered, but Power is pretty consensus the top player in this draft, and a #1 pick is very much expected to be a #1 d-man.

He's the #1 pick because it's a weak draft year at the top. Just because he's #1 this year doesn't mean he is expected to be a #1 defenseman anymore than a guy like Jamie Drysdale or Jake Sanderson would be.

At the end of the day, Power still has a number of warts in his game and he doesn't really have any elite traits to his game. I don't know how we've come to overlook all of this and act like Power is the most certain defensive prospect in the last 5 years.

 

[Blackjack]

It doesn't matter what you or I think, the question is if he beats the market. If his predictions, ordered every year, beat the NHL draft, then he is on to something even if that tool is not particularly helpful at picking an individual player. As I've said upthread, I would guess this beats the NHL draft in aggregate, but only very slightly, and again, that when something like this misses, they tend to miss bigger than NHL teams miss these days. It's not very useful to NHL teams except as a way of comparing players' numbers across leagues which can be quite a challenge given that there's ~10+ leagues represented in the first 3 rounds of the NHL draft. It can help you look at the players your team should want to draft.

I've soured a bit on Bader. His "star player" thresholds of 0.5 ppg for defensemen and 0.75 ppg for forwards seem totally arbitrary to me. He also kept insisting that Gusev was going to be like Panarin level or something, long after it was apparent that Gusev would never be anything more than a mediocre power play specialist at best.

Maybe his NHLe formula beats other public models for comparing feeder leagues and European leagues, and if that's the case, he certainly deserves some credit for that, but I don't put much stock in his opinions any more.

 

[Unknown Caller]

I've soured a bit on Bader. His "star player" thresholds of 0.5 ppg for defensemen and 0.75 ppg for forwards seem totally arbitrary to me. He also kept insisting that Gusev was going to be like Panarin level or something, long after it was apparent that Gusev would never be anything more than a mediocre power play specialist at best.

Maybe his NHLe formula beats other public models for comparing feeder leagues and European leagues, and if that's the case, he certainly deserves some credit for that, but I don't put much stock in his opinions any more.

Gusev wasn't a good look for him, but his model is usually very effective. If you look at guys like Point, Kucherov, DeBrincat, etc. he is ahead of the "industry" most of the time. He's never going to bat 1.000, but you need to look at his results compared to industry scouts. He is very good at limiting bad decisions and finding buried talent in my opinion.

 

[devilsblood]

He's the #1 pick because it's a weak draft year at the top. Just because he's #1 this year doesn't mean he is expected to be a #1 defenseman anymore than a guy like Jamie Drysdale or Jake Sanderson would be.

At the end of the day, Power still has a number of warts in his game and he doesn't really have any elite traits to his game. I don't know how we've come to overlook all of this and act like Power is the most certain defensive prospect in the last 5 years.

Like I said I think Dahlin would be the one guy who predraft was really thought to be a true #1.

The other guys? I think Power would rank right there with them. Or at least it wouldn't be crazy to say he's as good prospect predraft as any of them.

Now is Power a lock to be a #1, that is rarely the case with D-men, but it's not unrealistic to think that he will be.

 

[Unknown Caller]

Like I said I think Dahlin would be the one guy who predraft was really thought to be a true #1.

The other guys? I think Power would rank right there with them. Or at least it wouldn't be crazy to say he's as good prospect predraft as any of them.

Now is Power a lock to be a #1, that is rarely the case with D-men, but it's not unrealistic to think that he will be.

If Power were in the 2019 or 2020 draft he wouldn't sniff #1. He is not in a class above Seider, Byram, Drysdale, Sanderson, Broberg, Heiskenen, Makar, etc. He's around the same tier, maybe a little lower than most of those guys.

 

[Blackjack]

Gusev wasn't a good look for him, but his model is usually very effective. If you look at guys like Point, Kucherov, DeBrincat, etc. he is ahead of the "industry" most of the time. He's never going to bat 1.000, but you need to look at his results compared to industry scouts. He is very good at limiting bad decisions and finding buried talent in my opinion.

Point, Kucherov, and DeBrincat were all drafted and became high profile players before he started making predictions. You can't backtest a model like that, you can only grade it against real predictions it makes.

 

[Unknown Caller]

Point, Kucherov, and DeBrincat were all drafted and became high profile players before he started making predictions. You can't backtest a model like that, you can only grade it against real predictions it makes.

Not really, his model is a basic NHLe model that takes into account a player's profile (league, height, weight, etc.) and point production and then spits out a predicted outcome. He can 100% backtest the model and he has. All you need are a player's background and point totals as a prospect to get a draft year evaluation. You can look up tons of guys from the 90s in his model.

 

[devilsblood]

If Power were in the 2019 or 2020 draft he wouldn't sniff #1. He is not in a class above Seider, Byram, Drysdale, Sanderson, Broberg, Heiskenen, Makar, etc. He's around the same tier, maybe a little lower than most of those guys.

But if he is around the same tier then why is it crazy to think he could go #1 amongst them?

Even this year, I think he benefits from Hughes's injury. If Luke had finished the year strongly, perhaps went and had a big U-18 tourney, then maybe Luke would be getting the love. Or if Clarke was able to play a full OHL season. It's a tier and Power is at the top of the tier at the moment.

Which kind of backs up your point, but I think my point is more to temper the talk on either end of the extremes. Is he clearly the best guy of the past 5 years? No. Were guys like Makar clear #1's predraft who would def be ahead of Power this year? I'm not so sure about that.

 

[devilsblood]

Not really, his model is a basic NHLe model that takes into account a player's profile (league, height, weight, etc.) and point production and then spits out a predicted outcome. He can 100% backtest the model and he has. All you need are a player's background and point totals as a prospect to get a draft year evaluation. You can look up tons of guys from the 90s in his model.

I'm sure his model produced some hits from the past, but I imagine if you go through you'd find a fair amount of misses as well. Was he the guy who had Speers with the same projections as Rantanen?

 

[Unknown Caller]

But if he is around the same tier then why is it crazy to think he could go #1 amongst them.

Even this year, I think he benefits from Hughes's injury. If Luke had finished the year strongly, perhaps went and had a big U-18 tourney, then maybe Luke would be getting the love. Or if Clarke was able to play a full OHL season.

Which kind of backs up your point, but I think my point is more to temper the talk on either end of the extremes. Is he clearly the best guy of the past 5 years? No. Were guys like Makar clear #1's predraft who would def be ahead of Power this year? I'm not so sure about that.

Again, he is going #1 because there are no elite prospects in this draft. I'm not sure how this can be used as an evaluation metric against guys in much stronger draft years. Power would not go top 3 last year. He would not go top 3 in the Hughes/Kakko/Dach year. He would not go #1 (or probably not top 3) next year with Shane Wright coming out.

Also remember that Power is almost an entire year older than Luke Hughes and is closer in age to Drysdale and Sanderson.

 

[Blackjack]

Not really, his model is a basic NHLe model that takes into account a player's profile (league, height, weight, etc.) and point production and then spits out a predicted outcome. He can 100% backtest the model and he has. All you need are a player's background and point totals as a prospect to get a draft year evaluation. You can look up tons of guys from the 90s in his model.

But that doesn't tell you anything about how good the model is about making predictions. You can simply tweak the model until correlates well with successful NHL picks in the past, but that doesn't mean it will work going forward. This is a basic principal of modeling. Generally you do need to rely on historical data when creating the model, but it's unproven until it makes its own predictions, otherwise the model may just be fooled by statistical noise in the data.

 

[devilsblood]

Again, he is going #1 because there are no elite prospects in this draft. I'm not sure how this can be used as an evaluation metric against guys in much stronger draft years. Power would not go top 3 last year. He would not go top 3 in the Hughes/Kakko/Dach year. He would not go #1 (or probably not top 3) next year with Shane Wright coming out.

Also remember that Power is almost an entire year older than Luke Hughes and is closer in age to Drysdale and Sanderson.

Makar and Heiskanen went in a draft without elite prospects, and they didn't even go #1. Same for the guys from last year.

 

[Comparison Ford]

after reading a couple articles about Clarke’s character, i am thrilled we are most likely going to get one of him or Hughes

 

[Nocashstyle]

If Power were in the 2019 or 2020 draft he wouldn't sniff #1. He is not in a class above Seider, Byram, Drysdale, Sanderson, Broberg, Heiskenen, Makar, etc. He's around the same tier, maybe a little lower than most of those guys.

I get your point, but Seider wasn’t even in a class with Seider at the time of the draft. I don’t think Broberg should be included in that group either.

 

[nugg]

I can't believe you used 'star %' and 'ludicrous' in the same post not pertaining to each other.

Bader's SnakeOil% is 939% according to my formula. I saw his Framus score as low as 13 Grommets in some sites.

It's hard to believe someone could:
1. Quantify and distill the proper formula for someone being a 'star'.
2. Have sufficient data on these young players to make any meaningful prediction.
3. Do this during COVID, with even less data.

My biggest problem is slapping a percent sign behind an opinion just so others will swallow the hook.

I might put more stock in those "Eklund" ratings, e5

 

[Triumph]

But that doesn't tell you anything about how good the model is about making predictions. You can simply tweak the model until correlates well with successful NHL picks in the past, but that doesn't mean it will work going forward. This is a basic principal of modeling. Generally you do need to rely on historical data when creating the model, but it's unproven until it makes its own predictions, otherwise the model may just be fooled by statistical noise in the data.

There's not much reason to think that it's being fooled, though, unless there's been some massive change in either the NHL or the leagues that feed it. You are correct that the model is decidedly unproven.

 

[TrufleShufle]

I can't believe you used 'star %' and 'ludicrous' in the same post not pertaining to each other.

Bader's SnakeOil% is 939% according to my formula. I saw his Framus score as low as 13 Grommets in some sites.

It's hard to believe someone could:
1. Quantify and distill the proper formula for someone being a 'star'.
2. Have sufficient data on these young players to make any meaningful prediction.
3. Do this during COVID, with even less data.

My biggest problem is slapping a percent sign behind an opinion just so others will swallow the hook.

I'm working on my own currently, it is "PCD%." Right now everyone falls under "pretty cool dude" or "not cool, man." I hope to get more in depth, maybe add a, "he ain't a bad guy" or even "heard some stuff, but he's always been cool to me" category to get a bit more refined, then I'll start working on the % of it all.

 

[Blackjack]

There's not much reason to think that it's being fooled, though, unless there's been some massive change in either the NHL or the leagues that feed it. You are correct that the model is decidedly unproven.

Reasons that I'm skeptical:

1. His model makes an arbitrary cutoff for what constitutes a star: 0.45 ppg for a defenseman and 0.7 ppg for a forward (my numbers above were a bit off, I went back and checked.) Where did these numbers come from? Why are they significant? What if I tweak them up or down 0.05 ppg, do the model's results suddenly go down the toilet?

2. His method seems to be to assume that players that score a lot in their feeder leagues will generally score a lot in the NHL, and thus have a better chance at becoming a "star". That seems like an uncontroversial position, and the biggest insight, that short and Russian players are overlooked, is not exactly novel. I don't know about Kucherov, but I'm pretty sure I remember Point, DeBrincat, and some of the other guys he pumps as late round steals were seen as steals from the moment they were drafted.

As an aside, he's a big Marco Rossi booster and so am I. I don't think you need his fancy model to think that Rossi went too late in 2020, but if he ends up as a "star", I'll absolutely count it as a win for him.

3. He really hasn't provided evidence that having more stars on your team is the key to winning.

I think time will settle this one. If he ranks draft eligible prospects according to star probability, we can go back and check his list against who teams actually chose and who is right more often.

 

[Blackjack]

Okay sorry I didn't realize that we just had this discussion in this very thread.