Trollminator wrote:
Racket wrote:
Please take my advice, as someone with a PhD in math, that you should probably not repeat that line every again lol.
Anyways, you apparently can apply mean reversion (go figure) but it seems like a fundamental argument between polls having momentum versus random walks. If you accept that polls aren't random walks then mean reversion applies. Seems reasonable enough
This seems to be informative:
http://www.stat.columbia.edu/~gelman/research/published/psq_4021.pdfI should have replaced the word math with "numbers/data". Sorry to offend your superior credentials.
I know that you meant "numbers/data" when you said "math" and I said what I did regardless. You cannot blindly pull some model out of a book and apply to any data you happen across. There can be dozens of underlying conditions required of the data before analysis, usually first and foremost in regards to it being random or not.