Is 538.Com's Potential Explanation of the National-State Polling Gap Accurate?

[NOVA Green]

So, as many of us have been discussing on the various polling threads, there appears to be some interesting variance between state versus national polling models.

http://fivethirtyeight.com/features/election-update-the-blue-state-polling-abyss/


Nate Silver and 538 posted an article today discussing this in a bit of detail, and one of the arguments was that there is a major gap in polling of traditionally blue "non-battleground" states.

Do you believe that this explains the variance or not, and regardless please discuss.

I think it could explain some of the odd polling numbers we have seen this year, but does it explain why states like SC, GA, and TX look much closer than expected, or is it just another means of saying "we don't know, but maybe this is why".

[Fmr President & Senator Polnut]

Plausible, but unlikely.

Clinton's numbers will shift closer in line to a national trend toward the end. She might slightly under-perform Obama in a few places, but equally, we're still basing things off very few polls. If NM, ME and MN were being polled as frequently as VA or OH and the trend was there, then there might be a good discussion to have.

[NOVA Green]

So the article makes the case that Trump is significantly overperforming Obama '12 numbers in NY/NJ/CT in particular, which obviously contain a huge chunk of the national vote....

Now, I am bit skeptical about Trump's prospects versus Obama '12 numbers in WA, OR, CA, as well as in parts of the Upper-Midwest (MN, WI, MI), and suspect that the NE numbers won't trend significantly Republican (Excepting ME with minor swings in MA and CT).

So unless Trump is significantly overperforming Romney '12 in NY & NJ, what other "Blue States" are there to explain the overall National numbers versus much closer state numbers in Red States???

[Fmr President & Senator Polnut]

So the article makes the case that Trump is significantly overperforming Obama '12 numbers in NY/NJ/CT in particular, which obviously contain a huge chunk of the national vote....

Now, I am bit skeptical about Trump's prospects versus Obama '12 numbers in WA, OR, CA, as well as in parts of the Upper-Midwest (MN, WI, MI), and suspect that the NE numbers won't trend significantly Republican (Excepting ME with minor swings in MA and CT).

So unless Trump is significantly overperforming Romney '12 in NY & NJ, what other "Blue States" are there to explain the overall National numbers versus much closer state numbers in Red States???

I think it's whether the Dems might be maxed out in certain places and Trump is performing as well as Romney in places like IN, KY, WV etc...

But state polls are a better indicator, particularly with uncertain ballot placements for third parties.