By using this site, you agree to our Privacy Policy and our Terms of Use. Close
nuckles87 said:
gergroy said:
PDF said:
Obama has this locked up. What will be interesting is if Romney wins the popular vote.

I don't know, a lot of these polls out there are all within the margin of error.  It will come down to voter turnout.  I wouldn't say Obama has it locked up, but I would say he is the favorite to win.  


Margin of error only matters for individual polls. In polling composites, the margin of error is reduced SIGNIFICANTLY. Right now most swing state polls and national polls are moving in Obama's direction.

Not actually true.

Multiple polls signficiantly reduce the margin of error.... if they're looking at the same thing and their stnadads are fairly even.

All polls have their own models and often measure different things and also oddly, try and be as next to each other as possible because nobody likes risking being "too far out there."


Which is why composite polling has actually done a pretty shitty job at predicting vote totals.

It's just rarely come into play in elections because elections have been so far apart usually.


The 2008 presidental primary polls come to mind... and hell... the presidential election polls come to mind.

http://www.livescience.com/2183-presidential-polls-wrong.html

http://ns.umich.edu/new/releases/7067

Rasmussen was considered an outlier... yet won first place in accuracy the last couple times around.

 

Polls are fun.... but not remotely definitive.