Answer about poll methodology in response.

[Brian37]

I've been following 
http://www.realclearpolitics.com/epolls/...-5491.html

Which has a averaged out gap currently of 5.8 in favor of Clinton. 

Not just this election, but since Obama's first run against McCain. I don't trust polls as a long term predictor but as a short snapshot for a few days are week, referencing multiple polls then averaging them all out, it does give you a better indicator, all be it imperfect thus "margin of error" added as the disclaimer.

I am only asking about different methodology, because the following  website uses the same polls too. But in this Huffpo average of polls she is ahead by almost 8 points.


http://elections.huffingtonpost.com/poll...vs-clinton

Now again, in your answers I am strictly asking about methodology, because in the Huffpo polls they include Not just Clinton and Trump, but "undecided" and "other".

But they do reference the same polls as RealClear does in the first link.

What is the difference outside adding two more categories? And which do you think would have a better methodology?