« ALEA Home | Email msg. | Reply to msg. | Post new | Board info. Previous | Home | Next

Re: per silver, consensus of pollsters **

By: Cactus Flower in ALEA | Recommend this post (0)
Tue, 19 Jan 16 8:28 PM | 84 view(s)
Boardmark this board | The Trust Matrix
Msg. 17956 of 54959
(This msg. is a reply to 11192 by Cactus Flower)

Jump:
Jump to board:
Jump to msg. #

explanation of 332.




» You can also:
- - - - -
The above is a reply to the following message:
Re: per silver, consensus of pollsters **
By: Cactus Flower
in ALEA
Wed, 31 Oct 12 8:29 PM
Msg. 11192 of 54959

Hi j-t,

I have o winning florida. that's how i get my 332 (i think colorado, virginia and florida will all end up blue). but that may just be my polling bias!!

interesting that all the pollsters are almost exactly in alignment with one another.

http://fivethirtyeight.blogs.nytimes.com/2012/10/31/oct-30-what-state-polls-suggest-about-the-national-popular-vote/

in case you use rcp, they seem to have the broadest band for swinginess for each state. maybe up to 5 or 6%. they also seem to take rasmussen seriously, whereas i think the others maybe see a little polling bias there also!

but my review of 2008 at pollster seems to suggest that the swing state poll averages are accurate to +/- 1%. so i think rcp is too cautious. at >1.5% one day before election day, that state is very, very likely to go with the polling average. so anything above 2% now in a heavily polled state is a hill and anything above 3% is a mountain. barring catastrophe, there just isn't enough time for much of anything to change the result.

apparently the florida republicans are seriously worried by early voting and absentee ballots. https://twitter.com/EvanAxelbank/status/263497034353483776/photo/1


« ALEA Home | Email msg. | Reply to msg. | Post new | Board info. Previous | Home | Next