Articles and Analysis


Actual vs Predicted Electoral Vote Chart

Here is a first look at the Electoral Vote chart for final trend estimates and actual vote. Note that the state votes are NOT final and a few are very close while others have significant count left to go. Still, we gotta look, right? States are ordered by margin. Colors for actual vote are those of the states based on final pre-election trend.

At DTW. Will try one more post before next flight!



Talk about close as hell.


Your site did well in predicting the outcome. So did electoral vote dot com and 538 dot com. Thanks go to each of these fine web sites for their work.

Please answer a poll on single payer univeral health care at http://poll.democratz.org


KMart Dad:

Did Obama run a great campaign or what? If he does as good a job with the country, the next four years are looking up.



Rasmussen totally NAILED the final national vote, margin and final tallies; I am impressed. Gallup wasn't even close; this must be an embarrassment to them and they must go back to their models to figure out where they went wrong.

As for the state polling, definitely a mixed bag. Rasmussen was off on most of his battleground final releases; PPP I feel did really well in their final releases; Selzer was stellar, as usual; and Zogby was Zogby.



Oh, and Nate Silver is The Predictor. He got 49/50 states right (assuming NC stays Obama) in the presidential election but screwed up some of the senate projections (WTF happened in Alaska???)



PPP was outstanding in the final polls.

I also love that they clearly present the breakdown between early voters and those planning to vote on election day.

They correctly noted that CO and NV were both basically in the bag for Obama, in view of the heavy early voting.

I also like that they blog about their results.

Ras did really well nationally. PPP I think had them beat in the state polls, but they we're far off.

Mason-Dixon was too pessimistic for Obama.



The final calculations for popular vote was almost exact (estimate 52.3% to 46.2%) over at 538. The electoral count was just as exacting (at least until NC is called for Obama)



What's wrong with Alaska ? When people say they are tired of corruption and then vote for a convicted crook like Ted Stevens, then the people have only themselves to blame.

Can anybody help with the Oregon and Minnesota Senate races?



Thank you Mark and everybody else at Pollster.com.
You did allow to easily stay abreast of what was going on in this historic campaign.
It is thank to you and other similar outfits across the WWW that comments as those made by CNN analysts about the level of knowledgeability of voters showing at Grant Park last night were possible. I think that along those line an informed voter is the best type of voter possible.
Thank you!



Stevens' conviction was the death knell for Begich.

Steven's victory means that Sarah Palin will likely be the next senator from Alaska. Alaskins really like Sarah Palin. She is much more popular than Begich.

If Stevens had been acquitted... it would be unlikely the Senate would/could expel him from the Senate, but he is still tainted. And i the case the electorate would expell him: Begich wins.

Conversely, Stevens convicted means that the Senate boots him out and Sarah Palin replaces him... most likely with herself. But that only happens if Stevens beats Begich.

This kind of weird scenereo is very difficult for a poll to pick up on, no matter how good. Especially when the key event, Stevens' conviction, happened so close to election day.

I give the pollsters a pass on this one.



Great election. Congratulations everyone (you too Boom). Anyone know what happened with Prop 8 out in CA?



I haven't compared different pollsters, but just looking at the average predictions heading into yesterday, both here and on RCP, it looks like most of the state poll averages were pretty close to accurate (with a few exceptions: GA, PA, ND, etc.).

But the national averages were a bit high. I'm trying to wrap my head around that apparent contradiction.



Gallup likely would have been right if turnout was higher.

It looked like we'd get over 130M last night, but that looks well out of reach now.

There's about 118.7 million votes so far for Obama and McCain. The other vote got a little more than 1%, so we're probably at about 120 million total votes right now.



Prop 8, sadly, passed.

The Minnesota Senate race is headed for a recount (the loser has the right to demand one if it's less than 0.5%).

The Oregon Senate race is too close to call. At the moment, Gordon Smith is up by a few thousand, but the most uncounted votes seem to be in Benton, Washington, and Multnomah Counties, which could bode well for Jeff Merkley.



News of the day (polling-wise)

1) Youth turnout only +1 compared to 2004. I think this is probably why a few of the national polls were too high.
2) No Bradley effect (hopefully this unsupported theory is now put to rest for good)
3) State polls were generally right on at the end. The finals for PA, NM and NV underestimated BO support, while VA and NC seemed to overestimate it slightly.



About that youth turnout: from what I've heard, it was only +1% if you measure it as a percentage of the total electorate... but since overall turnout was higher, that would mean a lot more youngsters voted this year than in the past. It's just that they were then swamped by everyone else.



Just because the youth turnout may have only been +1 this election should lead people to undervalue the role they played in this election.

First, Obama did about 10 points better in the youth vote than Kerry (not looking at the numbers now, so please correct me). Second, assuming the conventional wisdom that Obama's organization and ground game made a huge difference, that advantage is partly attributable to the role that younger folk played as volunteers in Obama's campaign.



And actually, because the youth were overwhelmingly supporting Obama (by some estimates 2:1 or 3:1) it actually was worth 2-4% vs. parity.



Thanks Adam. I just heard on Fox that not all prop 8 votes have been counted, but that it looks like it will pass. It also looks like there will be a number of legal challenges to it.

Anyone else watching Fox? They're amazing. They're claiming that Obama's win isn't large enough to qualify as a mandate. It's funny how Bush felt he had enough of a mandate to do whatever he wanted, even though he lost the popular vote in 2000, and just narrowly won it in 2004!



Congrats to the Dem's and all the hard working Obama supporters... you did turn out the vote!

This election was won by the independents siding with Obama. Whether it was a vote against Bush, or a vote for change by those moderates, we'll see in the upcoming years.

Just remember, it was only a short while ago, that the Repubs controlled the legislative and executive branches of the federal government, only to be side-lined now. If the Democratic leaders govern ineffectively and one-sided over the next few years, I imagine the same fate may face them.

I'm off to begin planning for 2012... the $750 million will look paltry in that election. Can anyone say 1 billion? 2 billion? Sad isn't it?


Vicente Duque:

North Carolina is going to be an Obama Win.

According to INTRADE bets :

Prices of bets to win 100 dollars at INTRADE :

Betting your dollars for Obama, Prices :

North Carolina ………….. 95.2, McCain 1.5

Indiana …………………..98.5, McCain 1.0
Missouri …………………. 2.0, McCain 92.0

These are the only states posted at INTRADE.

Thanks again for your excellent information.



Vicente Duque



I wanted to thank you Mark and the rest of the people at Pollster.com for a great site and all the incredible work that you have done during this election.

Although I am glad about the outcome of this election (as I am an Obama supporter), I am sad that this country still seems to be dramatically divided (both politically and culturally). I am also sad to see that the divisive politics of fear and misinformation are still being used and that in this day and age still seem to work. I have problems understanding how people can be easily misled by special interests and corrupt politicians. I really hope that we, as a nation, overcome the hate and fear and take a different path in the 21st century.

Thanks again ... signing off,




Where's all the republicans? They should have a big "GOP" sign hanging around their necks.



Overall youth turnout may not have increased much, but young people aren't homogenous. The margin for Obama was massive in the youth vote, suggesting that:

1) Some young voters switched from Bush to Obama. I cant imagine that a huge number did, but this appears to be the consensus opinion.

2) A number of young Bush voters disengaged and a number of non-voters from the past elections engaged. This is the likeliest scenario in my book, especially given how uninspired left-leaning young people were by the Kerry and Gore campaigns.



Probably spoke too soon about the final vote. Looks like there are still a lot of votes out:


My preliminary national turnout rate for those eligible to vote is 62.6% or 133.3 million ballots cast. This number may yet rise further as absentee ballots arrive and provisional ballots are processed, particularly in some western states. Until these outstanding ballots are counted, I would like to provide a conservative estimate. This turnout rate would be the largest since the 62.8% of 1964. If we top that number, which we might, the next highest turnout rate would be 63.8% in 1960.

That's much better!




Sorry, but that is what would be completely wrong.

I'm from Germany, and if there's one lesson at least some people over here have learned, then it's NOT to discriminate at minorities - especially not when they are not in a position to effectively defend themselves.

Yes, the GOP might be completely wrong, politically, and they have suffered a smashing defeat last night, prima facie.

But it's only 52:47 in the popular vote, and therer are a lot of areas - the "battleground states" - where there is a split [maybe even a cleavage] right through the middle of the electorate.

America has seen at first sight large majorities for Democratic candidates before, and the last three times, 1992, 1976 and 1964 after a brief time there came a HARD return by a rise of VERY agressive conservative/reactionary movements (Gingrich et al., Reagan and Nixon [the latter built on Goldwaterites...]) in congress and/or as Presidents.

This is not the time for over-confidence or gloating.

Actually, there never is one.




"Conversely, Stevens convicted means that the Senate boots him out and Sarah Palin replaces him... most likely with herself. But that only happens if Stevens beats Begich"

I think that Palin cannot replace Stevens with herself under Alaskan law. She has to call an election within 90-days of a Senators death, resignation, or inability to fulfill the requirements of office. So it's very likely that, if he doesn't win the election outright, Begich would run again...perhaps against Palin.

Palin might do so, simply because that would give her a 6-year position...in Washington's spotlight. That would make her a media darling, although she's have to learn that Africa is not a country but a continent...and which countries belong in NAFTA (hint "North America Free Trade Agreement"). In addition, she'd be freed from the difficulties of the Alaska Executive Branch Ethics Rules. She could accept money for media appearances and her erudite (ghostwritten) writings. She would no longer be sniped at about her cronyism and nepotism in Alaska that could turn her out in two years.



"Rasmussen totally NAILED the final national vote, margin and final tallies; I am impressed. Gallup wasn't even close; this must be an embarrassment to them and they must go back to their models to figure out where they went wrong."

Actually the people that had the election pinpointed almost exactly was Pew Research. They were just as accurate as they were in 04. I think I know who to trust the next time around in 2012.


You forgot to mention the TruthIsAll Election Model, which projected an expected 365.3 electoral votes for Obama and a 53.1% vote share.

The EV was calculated as the sum of the individual state expected electoral votes:

Expected EV = prob(i)* EV(i),

where prob(i) is the probability of Obama winning state(i). The probability was based on the average of the last 2-3 national polls (3% MoE) and assuming that 60% of undecided voters would break for Obama.

The expected EV formula is very simple. It's just a basic probability calculation of the THEORETICAL expected value.

A full -blown search of billions of combination EV scenarios performed by another well-known model, is total overkill.

It is interesting to note that the Election Model Monte Carlo EV simulation of 5000 election trials produced a 365.8 EV for Obama - illustrating the Law of Large Numbers and TRUE CONVERGENCE TO THE MEAN.

Obama won ALL 5000 election trials therefore he had a 100% win probability.

But a post-election analysis based on 2004 returning voters and the 2008 National Exit Poll indicates that Obama did much better (2-4%) than the recorded vote indicates.



Post a comment

Please be patient while your comment posts - sometimes it takes a minute or two. To check your comment, please wait 60 seconds and click your browser's refresh button. Note that comments with three or more hyperlinks will be held for approval.