Prospect Info: All-Purpose Prospect Thread v17.0

[ello]

There is some value in Bader's model, as well as other models used for projecting prospects.

But models aren't everything. If a model can't accurately account for a prospect missing 80% of their season due to hip surgery, than that model has some very obvious flaws.

isn't the model indirectly taking injuries into account? Like Bader said it doesnt matter why the player had such a poor season (poor performance, mental health, injury): a bad year in such a crucial developmental season, regardless of cause, has really bad developmental ramifications.

I think Nazar couldve spent another year at college and been better off for it. Kid basically lost a whole year. We'll see how this year goes. If he produces at a high level in the AHL id say that would quell most of our concerns

 

[Kevin Musto]

How many times do I have to request that Bader posts are banned here?

 

[Hawkster]

 

[Northernhawk]

Excited to able to go see my first NCAA hockey game this upcoming October 18th

Minnesota @ UMD
4 Blackhawks prospects
Oliver Moore and Sam Rinzel vs Adam Gajan and Dominic James

Hoping Gajan starts!

 

[Crow]

isn't the model indirectly taking injuries into account? Like Bader said it doesnt matter why the player had such a poor season (poor performance, mental health, injury): a bad year in such a crucial developmental season, regardless of cause, has really bad developmental ramifications.

I dont know why someone would make the assumption that a bad season and an injured season are the same without some kind of evidence. Lumping them all together does nothing to prove that assumption.

 

[DisgruntledHawkFan]

Stat models don't take injury or circumstance into account because it's all numbers. It's how you get the Preds third line center being more valuable than Patrick Kane.

 

[statswatcher]

how long has bader been doing his model? what kind of predictive power has he been able to demonstrate? has he applied his model to historical draft classes based on contemporaneously available information and compared its predictions to what we know happened historically? can he even do an experiment like that given the data he feeds into this proprietary black box?

models are very very useful. they are the single most effective tool human beings have to build working understandings of a dynamic world with imperfect knowledge. that usefulness is directly proportional to the model's historical predictive power.

 

[belfour30]

how long has bader been doing his model? what kind of predictive power has he been able to demonstrate? has he applied his model to historical draft classes based on contemporaneously available information and compared its predictions to what we know happened historically? can he even do an experiment like that given the data he feeds into this proprietary black box?

models are very very useful. they are the single most effective tool human beings have to build working understandings of a dynamic world with imperfect knowledge. that usefulness is directly proportional to the model's historical predictive power.

I don't know how long Bader has been doing his model but I do know that he has retroactively gone back at least 30 years.

 

[statswatcher]

I don't know how long Bader has been doing his model but I do know that he has retroactively gone back at least 30 years.

i suppose my question is what does that mean, "gone back"? maybe he goes into more depth on his methodology somewheres, but i honestly can't be bothered to look, and i won't really be satisfied without being able to look at his formulas and see exactly what he's doing.

 

[belfour30]

i suppose my question is what does that mean, "gone back"? maybe he goes into more depth on his methodology somewheres, but i honestly can't be bothered to look, and i won't really be satisfied without being able to look at his formulas and see exactly what he's doing.

He takes data from ~1990 onward and puts them into his model. Honestly we know what the guys from 30 years ago turn into. It's just NHLe.

 

[statswatcher]

He takes data from ~1990 onward and puts them into his model. Honestly we know what the guys from 30 years ago turn into. It's just NHLe.

the problem in my mind is that in order to keep the model consistent, he's limited to the kinds of prospect statistics which were available 30 years ago. no possession, you can totally forget shot quality. be shocked if you could even find consistent time on ice data for prospects from 30 years ago depending on the league. what's he doing, plugging in points and plus/minus and calling it a day?

 

[GIADF]

Sounds like a bunch of made-up BS he's passing off as fancy stats.

 

[statswatcher]

Sounds like a bunch of made-up BS he's passing off as fancy stats.

to me this kind of stats influencer, throw some unspecified data into a spreadsheet, make a slick infographic to visualize what it spits out, and charge people for the privilege of checking your work is highly irritating, and a large part of why things like 'stats' and 'numbers' are looked at with suspicion by the average fan. all these guys give you is a conclusion. you get none of the value of seeing how that conclusion follows from real data and well reasoned inference.

 

[GIADF]

to me this kind of stats influencer, throw some unspecified data into a spreadsheet, make a slick infographic to visualize what it spits out, and charge people for the privilege of checking your work is highly irritating, and a large part of why things like 'stats' and 'numbers' are looked at with suspicion by the average fan. all these guys give you is a conclusion. you get none of the value of seeing how that conclusion follows from real data and well reasoned inference.

The only thing you forgot to mention is the temper-tantrum he'll throw when being questioned.

 

[Kevin Musto]

Clownshoe's model is literally just NHLe.

There are no fancy stats.

It's just point totals, adjusted to project equivalency in NHL scoring.

So if I say statistically AHL production translates to 0.5 in NHL equivalency, a player who scored 82 points in 82 AHL games would be projected to score 41 points in 82 NHL games.

That's what his model is.