Toggle sidebar

Prospect Info: General Discussion of Prospects

  • Xenforo Cloud will be upgrading us to version 2.3.5 on March 3rd at 12 AM GMT. This version has increased stability and fixes several bugs. We expect downtime for the duration of the update. The admin team will continue to work on existing issues, templates and upgrade all necessary available addons to minimize impact of this new version. Click Here for Updates
Ace said:
I sometimes forget there is an entire contingent of Russian prospects out there for us. Poltapov, Neuchev, Kisakov, Sardarian and Novikov have just been out of sight out of mind. That’s a lot of guys. Kisakov we will get to see this year now so that’s exciting.
Click to expand...
Neuchev seems promising and poltapov looks like he could turn into an effective bottom sixer
 
Yes. But I guess he gets credit for noticing the Sabres had 11 picks and 3 of them were in the first round? Does he have a patreon I can sign up for?
 
  • Like
Reactions: jmelm and SECRET SQUIRREL
Ace said:
Yes. But I guess he gets credit for noticing the Sabres had 11 picks and 3 of them were in the first round? Does he have a patreon I can sign up for?
Click to expand...

He accounts for that by grading a Team's picks against what the picks 'should' have been.

I'll say his evaluation metric is not very good for efficiency. He seems to be giving teams a C if they exceed their draft position EXP NHLer by < 0.75 (this is a bit of a guess on my part). A raw value isn't the correct metric. It should be a % of their EXP NHLer.

For example Florida's draft picks should only yield 0.76 NHLers according to his model, Florida drafted 1.35 NHLers; to me that's pretty good, but only gets a C value from Bader. They didn't have much but they were extremely efficient.

Florida exceeded expectations by 77%, Buffalo only exceeded expectations by 36%. Why does Buffalo get an A+? Just because they picked 4 EXP NHLer vs 2.95? That's good, but not incredibly efficient.

Additionally I'm a little curious how most of the teams exceeded their expected pick values? Is he scoring this as a strong draft? May be true, i'm not sure.
 
SnuggaRUDE said:
He accounts for that by grading a Team's picks against what the picks 'should' have been.

I'll say his evaluation metric is not very good for efficiency. He seems to be giving teams a C if they exceed their draft position EXP NHLer by < 0.75 (this is a bit of a guess on my part). A raw value isn't the correct metric. It should be a % of their EXP NHLer.

For example Florida's draft picks should only yield 0.76 NHLers according to his model, Florida drafted 1.35 NHLers; to me that's pretty good, but only gets a C value from Bader. They didn't have much but they were extremely efficient.

Florida exceeded expectations by 77%, Buffalo only exceeded expectations by 36%. Why does Buffalo get an A+? Just because they picked 4 EXP NHLer vs 2.95? That's good, but not incredibly efficient.

Additionally I'm a little curious how most of the teams exceeded their expected pick values? Is he scoring this as a strong draft? May be true, i'm not sure.
Click to expand...
What are the "expected" based on?
 
SnuggaRUDE said:
I'm not intimately familiar with Bader's work, but I assume it's a historical look at the likelihood of an NHLer at each pick.
Click to expand...
Hmmm. Then how would each pick grade out if they're just picks in spots? I'm guessing it's based on his (or maybe some consensus) rankings veiled in a bunch of numbers to look scientific.
 
Matt Ress said:
Hmmm. Then how would each pick grade out if they're just picks in spots? I'm guessing it's based on his (or maybe some consensus) rankings veiled in a bunch of numbers to look scientific.
Click to expand...

Well there's 2 things:

1. The likelihood of an NHL from each pick, based on historical data
2. The likelihood of a drafted player becoming an NHL, based on his model

It is indisputably scientific. You may not believe it's accurate.
 
SnuggaRUDE said:
Well there's 2 things:

1. The likelihood of an NHL from each pick, based on historical data
2. The likelihood of a drafted player becoming an NHL, based on his model

It is indisputably scientific. You may not believe it's accurate.
Click to expand...
Sure I guess you can call his model scientific, whatever that model may be. But building out more data on top of known imperfect and highly debatable data is the opposite of scientific. That's not how science works and it has nothing to do with what I believe about it.

Whatever. I like our draft and I think we did great in the 1st round
 
Matt Ress said:
Sure I guess you can call his model scientific, whatever that model may be. But building out more data on top of known imperfect and highly debatable data is the opposite of scientific. That's not how science works and it has nothing to do with what I believe about it.

Whatever. I like our draft and I think we did great in the 1st round
Click to expand...

You're entitled to your opinion but I believe your understanding what constitutes science is inaccurate.
 
SnuggaRUDE said:
You're entitled to your opinion but I believe your understanding what constitutes science is inaccurate.
Click to expand...
It's not that his model isn't science - it's that his model is flawed in the way it's designed.

I've gone over this before, but there are numerous problems, such as:

- Weighting Devon Levi being drafted out of the CCHL more than his Division 1 .952 sv% in his D+2 year. Bader still considers him a low probability prospect - primarily due to the CCHL + his size. ()

- Considering Peterka a low probability prospect because he was drafted out of the DEL - and that his prospect profile doesn't match up with historical stars - which is mostly weighting his performance in the DEL in his DY and D+1 year, and not his near-PPG pace in the AHL in his D+2 year. ()

- Not separating the USHL/USDP stats until recently (he didn't know they were combined in EliteProspects data)

Science is fine - but when you are producing data with a model that has many flaws and spitting out results that are very likely not true at this point - it's hard to take it seriously. I do think there's something to be said for statistical benchmarks that players should be at/around in their DY and D+1 year, and if they're not - they're statistically unlikely to have success. In which case, having a back-tested model to show that would be valuable. However, Bader's model has large blind spots in a number of areas. The model may be more valuable to compare players with a large sample size (i.e. players drafted out of the CHL, play D+1 and D+2 in the CHL) but players with a more unique path to the NHL (such as coming from a league that doesn't produce a high number of NHL players) - the model is likely to give you inaccurate data because the weight of that is too heavy in his model IMO.
 
  • Like
  • Love
Reactions: SECRET SQUIRREL, sabremike, Weltschmerz and 2 others
SnuggaRUDE said:
Well there's 2 things:

1. The likelihood of an NHL from each pick, based on historical data
2. The likelihood of a drafted player becoming an NHL, based on his model

It is indisputably scientific. You may not believe it's accurate.
Click to expand...

I wouldn't call it scientific. Objective might be the best word to use.
 
His Levi explanation demonstrates the downside of analyzing players this way. He's absolutely right that given the raw data, a player X who is in Levi's shoes is already an outlier, and that the few who have come before weren't all that great. It underrates how exceptional the actual accomplishment was and what it portends for Levi in favor of past players who aren't necessarily good comarables in situation despite similar data.
 
  • Like
Reactions: sabremike
RefsIdeas said:
- Weighting Devon Levi being drafted out of the CCHL more than his Division 1 .952 sv% in his D+2 year. Bader still considers him a low probability prospect - primarily due to the CCHL + his size. ()

- Considering Peterka a low probability prospect because he was drafted out of the DEL - and that his prospect profile doesn't match up with historical stars - which is mostly weighting his performance in the DEL in his DY and D+1 year, and not his near-PPG pace in the AHL in his D+2 year. ()

- Not separating the USHL/USDP stats until recently (he didn't know they were combined in EliteProspects data)

Science is fine - but when you are producing data with a model that has many flaws and spitting out results that are very likely not true at this point - it's hard to take it seriously. I do think there's something to be said for statistical benchmarks that players should be at/around in their DY and D+1 year, and if they're not - they're statistically unlikely to have success. In which case, having a back-tested model to show that would be valuable. However, Bader's model has large blind spots in a number of areas. The model may be more valuable to compare players with a large sample size (i.e. players drafted out of the CHL, play D+1 and D+2 in the CHL) but players with a more unique path to the NHL (such as coming from a league that doesn't produce a high number of NHL players) - the model is likely to give you inaccurate data because the weight of that is too heavy in his model IMO.
Click to expand...


Couple issues here might be the hangups.

1. I believe his method is scientific, not flawless. Science, particularly modeling uncertainty is an inherently iterative process.
2. We can't know if his modeling for Levi or Peterka is inaccurate yet, since they aren't NHLers.
3. Knowing a model is uncertain doesn't make it unscientific or even flawed. It just makes it less consistent. The error bars on players with unusual development paths are larger.

Bader's model, like virtually everyone else who's using some variation on NHLe + Age to project development are using models back checked against current players. They're lagging indicators of trends in the game and development systems.

It's very reasonable to say that players with Peterka's development pattern haven't seen much success at the NHL level; they haven't. Given the paucity of comparables Peterka if he makes the NHL, which I suspect he will, will greatly influence the modeling of players DEL->AHL->AHL.

I don't see anything which suggests the D / D+1 are weighed heavier than D+2. It's not part of your links, if it's posted elsewhere or common knowledge I'd understand.
 
How many worthy top 6 forwards will we have in 3 years?

Thompson
Tuch
Cozens
2 of Krebs, Quinn, Paterka
2 of Rosen, Kulich, Savoie, Ostlund?

Skinny? Asplund? Olofsson? Mitts?

Possibilities: Weissbach, Nadeau, Rousek, Bloom, Paltapov, Neuchev, Kozak, Kisakov
 
Matt Ress said:
How many worthy top 6 forwards will we have in 3 years?

Thompson
Tuch
Cozens
2 of Krebs, Quinn, Paterka
2 of Rosen, Kulich, Savoie, Ostlund?

Skinny? Asplund? Olofsson? Mitts?

Possibilities: Weissbach, Nadeau, Rousek, Bloom, Paltapov, Neuchev, Kozak, Kisakov
Click to expand...

I'd say it's entirely possible those numbers turn to 3.

No secret I'd still want Olofsson on the team.
 
  • Like
Reactions: Matt Ress
Next offseason I'd like to trade one of our young forwards for Andrew Peeke. Columbus is stocked with young RHD.
 
Matt Ress said:
How many worthy top 6 forwards will we have in 3 years?

Thompson
Tuch
Cozens
2 of Krebs, Quinn, Paterka
2 of Rosen, Kulich, Savoie, Ostlund?

Skinny? Asplund? Olofsson? Mitts?

Possibilities: Weissbach, Nadeau, Rousek, Bloom, Paltapov, Neuchev, Kozak, Kisakov
Click to expand...

Thompson, Tuch, Cozens - yes
2 of Krebs/Quinn/Peterka - yes
2 of Rosen/Kulich/Savoie/Ostlund - no

Even if they're going to eventually end up there they're just too young. Not many 21 year old 2nd liners in the NHL.

Skinner - no, he'll be 33
Asplund - no, he's an NHLer but not a top 6 player
Olofsson - no, he'll be 30
Mitts - I dunno, he might be one of those middle 6 players who can slot up and down a lineup.

The HMs: I don't think any of them will be ready for that kind of deployment in 3 years.
 
tsujimoto74 said:
Novikov could definitely become a good player. He did pretty well for himself last season playing in the KHL (and holding his own) as an 18yo.
Click to expand...
100%. I don't think he's being talked about enough. Him playing top pairing in the KHL at 19 is a big deal. And the guy is a monster out there, he's only listed at 6'4 but plays like he's 8 feet tall. Just a bull amongst sheep. I really hope he comes over next year.
 
  • Like
Reactions: SECRET SQUIRREL
In 24/25, four 2nd lines, order them however you want. So, twelve top 6 forwards. It was a trick question.

Peterka - Krebs - Savoie
Kulich - Cozens - Tuch
Skinner - Mittelstadt - Quinn
Asplund - Thompson - Olofsson
 
You must log in or register to reply here.

Latest posts

Ad

Upcoming events

Ad

Top Bottom