Msfann":34xzwghf said:
gaucho":34xzwghf said:
Hi Again Guys -
Football Outsiders promised me an article on the "morning body clock" issue when the schedule comes out later this month (hopefully they keep their promise). Anyways, below is a draft of what I'm going to send. I figured I'd post it here to see if anyone has any feedback, and also to give a shoutout to AgentDib whose statistic I used.
http://webpages.sou.edu/~stonelakb/math/pdf/Schedule Release.pdf
That's pretty cool, I hope that gets people talking and gets the nfl to listen. :th2thumbs:
Not to be that ninny, but anyone with a passing knowledge of data visualization is going to spend a half second looking at the y axis on the "home team win" graph and recognize the cooking of the books that's going on in here.
What the author is trying to get away with is even an example used in Darrell Huff's bestseller from the 1950s titled
How To Lie With Statistics.
As I tell the students I work with, if you find yourself doing stuff like that you need to slowly back away from your entire research project, and likely start the whole thing over again with an eye toward curiosity rather than advocacy so that you can rethink the other hundreds of small decisions you almost definitely made along the way to get to your desired conclusion. Even better yet, think about studying something you can be more dispassionate about, so that you don't have to constantly fight the urge to try to just confirm your priors.
For statistics, it's almost always a LACK of curiosity which kills the cat, IMO.