Thread: Justin Schultz
View Single Post
12-18-2012, 10:37 PM
Registered User
Fourier's Avatar
Join Date: Dec 2006
Location: Waterloo Ontario
Country: Canada
Posts: 10,653
vCash: 500
Originally Posted by Hardyvan123 View Post
26 games is not enough context. There is alot of variance that can happen in 26 games compared to a whole season. If you don't understand the concept of regressing to the norm then look it up.

Over 80 games, over several seasons we see what a players norm is.

I would be curious to know what his shooting % was like in the last 2 years in a weaker league.

I looked it up 16 goals on 110 shots is 14.54 (37 games)

The season before is 18 goals on 108 shots which is 16.666 (43 games)

In his rookie year in the WCHA it was 6 goals on 61 shots good for 9.83

Compared to Smith from Detroit at the same school his shooting % is better but it's still a long way to translate that into the NHL.
Based on your own words:

Originally Posted by Hardyvan123 View Post
I'm not spending much time on this thread because math really bores me and I don't understand all the technical mumbo jumbo (more due to a lack of interest than anything else) but I do understand the concepts.
I am going to guess that it might be you who does not understand the concept of "regressing to the norm". (By the way, the reason I use quotes is because the actual term is regressing to the mean. In mathematics a typically norm is an abstract length function defined on a vector space over either real or complex scalars.

The concept you are quoting is repeatedly butchered on these boards. In the context in which you have used it it does virtually nothing for your argument. Let me give you an extreme example of why your previous statement suggesting that "all players regress to the norm" is dubious. Since you have not made a case that this principle only applies to SH% I can assume it also applies to goal scoring. The mean number of goals per player is about 10. If I buy what you wrote then I can conclude the Steven Stamkos should be expected to regress to that level in the near future. after all you have rejected the idea that it is possible that elite skill might lead to statistical outliers.

Fourier is offline   Reply With Quote