Big Lizards:Blog:Entry “Some Interesting Polling Figures and Map Games”

« Marriage - a Fundamental Liberty? | Main | What to Do About Criminally Convicted Sen. Teddy? »

October 27, 2008

Some Interesting Polling Figures and Map Games

Hatched by Dafydd

Currently -- before most of the October 26th polling trickles in -- Barack H. Obama is ahead in the Real Clear Politics average by 7.6%; but that includes some whoppers (in both senses of the word) from several days ago, polls of dubious character: the Newsweek poll (Obama +12) and the CBS/New York Times poll (Obama +13). Both are outliers by far; no other poll shows a bigger spread than 8 points, except for the Gallup "expanded" poll -- Obama +9 -- which is essentially a poll of registered voters, not likelies, and which I completely discount... hence do not count.

Taking only the most recent polls that include October 25th or later, Obama's lead drops to 6.2% [correction, now down to 5.3% with a couple more polls].

But I suspect that any voter who is still undecided on election day will vote for John S. McCain: Obama is the riskier candidate of "hopey changitude," as Beldar puts it; and those who are hesitating are likely those who kinda sorta want to vote for Obama but just aren't sure he's up to the job, having virtually no resume at all. In any event, if we assign all the undecideds to McCain, that tells us the best McCain can get without having to pry Obama supporters away from the One They Have Been Waiting For.

Going through the recent polls and assigning all the undecideds to McCain gives us the following numbers: Barack Obama, 50.2%; John McCain, 49.8%... Obama leads by 0.4%.

Of course, there is a certain built-in bias towards Democrats in polling; it stems from several sources:

Exaggeration of probable Democratic turnout and a corresponding minimization of Republican turnout;

The "self-selection" fallacy, wherein Democrats are more willing to cooperate with pollsters than Republicans;

The "weekend polling" fallacy;

The "PC effect." I don't believe much in the pure Bradley effect -- voters saying they will vote for the black candidate but really voting for the white candidate, due to racism; but there is clearly a tendency for respondents to falsely tell pollsters they will vote the "politically correct" way, then vote the opposite in the privacy of the voting booth. This effect is especially pronounced when during a concerted campaign to paint anyone who doesn't vote for Obama as a "racist."

Given all this, depending on how the undecideds break, it's entirely possible that McCain would actually be ahead right now in some hypothetically perfect poll. But even if Obama would still lead, it's not by very much... a couple of percentage points at worst.

In the meanwhile, it's good to remember that this is not a single election but 51 separate elections (50 states plus the District of Columbia) that will determine the electoral vote. As I've noted in the past, Real Clear Politics has a facility where one can take an Electoral College map and reassign states at will, to explore possible routes to la Casablanca.

I believe I already mentioned one such scenario: If McCain ends up winning Colorado, Ohio, and Virginia, where he is currently running slightly behind in state polling, then he will almost certainly win every state that is currently a toss up as well (especially since every toss-up state is a state that George Bush won in 2000 and 2004); that would give McCain 274 electoral votes, and he would win with four points to spare.

But here is another route to the White House... even if McCain loses Colorado:

First, read this blogpost on Virginia Virtucon about a Democratic pollster's assessment of the real status of several battleground states... as opposed to the probably biased state polls that are reported on RCP (hat tip to "Radioblogger" Duane Patterson on Hugh Hewitt's blog). The female Democratic pollster says:

[T]he results of their polling lead her to believe that McCain will definitely win FL, OH, NC, MO and NV. She says Obama definitely wins New Mexico. She said that Colorado and New Hampshire were absolute dead heats. She said she thinks there is a 55% chance Obama holds on in Pennsylvania and a 75% chance McCain wins Virginia....
Anyway, her companies conclusion is that the election will come down to Colorado, New Hampshire and the Republican leaning district in Maine, which in her opinion might very well decide the Presidency.

Let's take her at her word; here is my alternative scenario. I had forgotten that Maine is one of only two states (I believe) that split their electoral votes. Assume McCain wins all the toss ups, and that he wins Ohio and Virginia but loses the Rocky Mountain state. That gives him 265, Obama 273. But now, if the pollster in the article above is correct, McCain could win New Hampshire.

The only poll showing Obama way ahead in NH is the Boston Globe's, one of the most notoriously biased polls around. If Rasmussen (Obama +4) is more correct there -- or even Rasmussen averaged with the Concord Monitor (Obama +7), the only two recent polls besides the Globe's -- then McCain is only 4 or 5 points down in that state... or 3 points if McCain gets the undecided vote.

If McCain wins New Hampshire, that makes it a 269-269 tie. But McCain has a very good shot at winning one of Maine's electoral votes, because they split: There is a Republican-leaning district that contributes one of the four votes.

If that happens, then even with Obama taking Colorado, it's McCain 270, Obama 268... and Keith Olbermann's head explodes like the overripe pumpkin it actually resembles, inside and out.

So as I have previously said, don't give up the ship until you see the whites of their eyes; get out and vote -- and get 3-5 friends to get out and vote, too!

Hatched by Dafydd on this day, October 27, 2008, at the time of 7:11 AM

Trackback Pings

TrackBack URL for this hissing: http://biglizards.net/mt3.36/earendiltrack.cgi/3305

Comments

The following hissed in response by: ManlyDad

Interesting analysis (and encouraging, thanks--something in short supply these days). Minor note:
50.2 - 49.8 = 0.4% lead.

The above hissed in response by: ManlyDad at October 27, 2008 9:39 AM

The following hissed in response by: ~brb

But... but... but according to Jonathan Alter in this week's Newsweek (http://www.newsweek.com/id/165657), if The Unabama does not win it can only be because of racism!

Nice impartial reporting there, Newsweek.

The above hissed in response by: ~brb at October 27, 2008 12:34 PM

The following hissed in response by: LarryD

Oh, don't forget this part:

Anyway, her companies conclusion is that the election will come down to Colorado, New Hampshire and the Republican leaning district in Maine, which in her opinion might very well decide the Presidency (apparently the district in Nebraska that Obama thought he might be able to get is now off the table). She said she has very little doubt that the public polling is part of a “concerted voter suppression effort” by the MSM.

The above hissed in response by: LarryD at October 27, 2008 1:17 PM

The following hissed in response by: Dafydd ab Hugh

ManlyDad: