Let's unskew the polls!

[Boston Bread]

While poll averages are generally reliable, in some years they have consistently underestimated one party's performance in an election. What will be the total poll bias in 2014, comparing aggregate poll averages with election results? Keep in mind that this is across the board, not in any particular race or pollster, and that by overestimation I mean how much the party underperforms in the election compared to their margin in polling average.
Edit: To be clear, a GOP overestimation of 3 points is like 2012, ie. they did worse than expected

[IceSpear]

So as I stated in the other thread, the polls haven't been meaningfully biased in favor of Democrats since 2002. That seems like a trend to me.

I'm gonna say GOP overestimated by 1-3, but I accidentally voted Dem overestimated 1-3.

[Boston Bread]

So as I stated in the other thread, the polls haven't been meaningfully biased in favor of Democrats seem 2002. That seems like a trend to me.

I'm gonna say GOP overestimated by 1-3, but I accidentally voted Dem overestimated 1-3.

Wording a poll like this isn't easy
I was trying to keep it consistent with the article, ie. a GOP bias being a bad thing for the GOP
Anyways, voted GOP overestimated by 1-3 but I'm hoping for a 1998 style long-shot of course. You can't really rely on pollsters underscoring your side (unless you're a D-HI) but there's some anecdotal evidence from recent elections at least.

[○∙◄☻¥tπ[╪AV┼cVê└]

I remember the 1998 election. Republicans were all giddy and had all these victory parties planned. D'Amato got stuck on the elevator on the way to his victory party and lost by 11 points.

[Maxwell]

I'll go ahead and say it'll be pretty neutral. Pollsters have seemed to learn their lesson about polling hispanics, though not all of them have learned their lesson about polling blacks.

Worst case? R+0.9, just like 2010