How accurate were the polls?

[pghin08]

I know it's a bit early to do this, but curiosity was getting the best of me. What I'll be using to compare with the results is the Real Clear Politics "poll of polls" so to speak. Let's go through the toss-up states:

Ohio:
RCP Average= Obama +2.9%
Real Results= Obama +1.9%

Florida:
RCP Average= Romney +1.5%
Real Results= Obama +0.5%

Virginia:
RCP Average= Obama +0.3%
Real Results= Obama +2.9%

New Hampshire:
RCP Average= Obama +2.0%
Real Results= Obama +5.8%

North Carolina:
RCP Average= Romney +3.0%
Real Results= Romney +2.2%

Michigan:
RCP Average= Obama +4.0%
Real Results= Obama +8.3%

Pennsylvania:
RCP Average= Obama +3.8%
Real Results= Obama +5.1%

Wisconsin:
RCP Average= Obama +4.2%
Real Results= Obama +6.5%

Iowa:
RCP Average= Obama +2.4%
Real Results= Obama +5.6%

Colorado:
RCP Average= Obama +1.5%
Real Results= Obama +4.7%

Nevada:
RCP Average= Obama +2.8%
Real Results= Obama +6.6%


This brings me to two conclusions:

1. Polls still leave a lot to be desired, and I still put no credence in them.
2. While the right lambasted them as being liberal-biased, in all cases but one, they actually underestimated Obama and overestimated Romney.

[JustAlex]

I'll tell you who is smiling today.....Nate Silver.

He gave Obama a 92% Chance of winning a full day before the election and he said we would come up with 314 electoral votes.

This is why I LOVE math, science, probabilities, and statistics.....they are NOT biased, they do NOT have an agenda, they only present data, you can take them or leave them, but you can't ignore them.

[Wickabee]

I think conclusion #2 is the most telling.

[habsheaven]

I think the polls under-estimated the election day ground game of Obama's campaign. And to be fair, it is really hard to reflect that in a poll.

[MadMan1978]

Kurt I have been saying all week you need to look what numbers are and what they are not...

Many Polls DID NOT have a large Minority polling data...In fact Some were 80% white.