MortiestOfMortys
Registered User
It seems like just taking raw data can skew everything. There are many cases, particularly in the QMJHL, where players post huge scoring stats and never have an impact in the NHL.
You’re right, and that’s what the model does. The Q is a high-scoring league, and it’s usually 20/21-year olds putting up those gawdy numbers. So by using scoring translation factors, and adjusting for age and era, you can control for those effects. In order to compare across leagues, you have to normalize your data somehow, and that’s how he does it (in a very broad sense).
What picks by Dallas are making them stick out like a sore thumb?
Harley is a good prospect for sure, but he went too early though. Their next 3 picks were real bad, which tied a boat anchor to their score and pushed it overboard (since it’s an average).