View Single Post
Old
04-13-2014, 12:11 PM
  #121
shinchanyo
Registered User
 
shinchanyo's Avatar
 
Join Date: Jan 2013
Location: Here
Posts: 2,741
vCash: 500
Quote:
Originally Posted by aufheben View Post
That's fine, my point was that ranking a tampered stat against untampered stats produces a meaningless number. It's a fallacy. Although I'm sure if it was -31- and not me who pointed it out, they'd all be singing his praises.
Meaningless implies that no change would occur in the data. It is meaningful in that regard. Whether it's accurate to make an analysis based on removing that segment is obviously the argument you're making though.

But snow is saying the former and he's right. Some teams do certainly face more adversity than others. Snow thinks the NYR started the season facing enough adversity to where it was abnormal compared to the rest of the entire league. If it is true that the NYR truly had an abnormal amount of adversity than in order to get an accurate gauge of the team it would make sense to remove a bit.

You may disagree that it was abnormal but it's on you then to provide evidence to show it. If you want to look solely at the NYR I think it's more than fair to remove the early games. If you want to compare to the rest of the league then I think it is incorrect to remove the sample from the NYR and not from anyone else.

Take the 10 worst games away from each team and then make your comparison. If the NYR truly had an abnormal amount of adversity then the adjustment made to the NYR will be MUCH higher than the adjusment made to other teams.


Last edited by shinchanyo: 04-13-2014 at 12:20 PM.
shinchanyo is offline   Reply With Quote