View Single Post
Old 12-03-2018, 06:40 PM   #121
Jay Random
Franchise Player
 
Join Date: Aug 2005
Exp:
Default

Quote:
Originally Posted by squiggs96 View Post
Is 6 of 10 statistically significant enough to put any link between those stats and if a team is good or not? I, personally, don't think it is. 6 of 10 is 1 better than 50/50, so it may as well be a coin flip.
Bingo's method picked from 31 teams, not 20. A given team has only a 32.26% chance of being in the top 10 by a given measurement. If you pick by coin flip or any other random method, you should expect to get 3.226 of the top 10 teams right.

So yes, it's a much stronger correlation than you assume. And that's with the limitations imposed by small sample size, since standings this early in the season are not always reflective of a team's overall quality. A five-game winning or losing streak has a much larger effect in a 25-game sample than in an 82-game season.
__________________
WARNING: The preceding message may not have been processed in a sarcasm-free facility.
Jay Random is offline   Reply With Quote