Quote:
Originally Posted by Double Down
Dumbish question re: statistics/polling:
Taking a look at Hillary's firewalls of CO, NM, MI, MN, WI, PA, VA, ME, she is on average an 80% to win each individual state according to 538.
If we don't take correlation into account for the sake of this argument, aren't the odds extremely against her to win them all? Going 8 for 8 on 80% shots is (.8)^8= 16.8%. So shouldn't it be expected for at least one or two of these to go red?
Even if we do factor correlation back in (and I have no idea how to do that accurately), it still doesn't seem like it would jump to over 50%.
I could be totally wrong/not taking certain things into consideration/talking out of my ass, etc.
Thoughts?
there's either large poll fail across the board in drumpf's favour due to shy racists or unexpectedly huge white turnout or something like that which none of the pollsters managed to capture, in which case approximately every one of those states you listed is in play
or there isn't, and approximately none of them are in play
therefore she doesn't need to bink that 80/20 over and over and over in order to win all those states (which as you note she'd be a huge dog to do), she needs to bink it just once
(imo there isn't 20% chance of drumpf-friendly poll fail but w/e ill use the 538 number)