Oct 25 2008
Bi-Poller
Jay Cost posted a good tutorial on the mathematics of polls, but left one case out where statistical models start to show two different possible outcomes. These are called bimodal results, where there is no consensus on a single result, but actually two highly probably results are possible. A graph of a bimodal statistical model is shown below for the duration periods between eruptions of the Old Faithful Geyser in Yellowstone National Park:
As can be seen there are two basic periods between eruptions which drive the timing of old faithful. This doesn’t really apply to polls but it is a good example of bimodal data sets.
If one looks at the polls in the election they are showing a two clusters of results. One cluster shows a runaway race for Obama, the other a very possible close win for McCain. For example, let’s look at the Polls in Ohio. We have 8 polls now on the RCP list today (which Jay Cost notes all RCP does is average, no real analysis and processing).
4 polls cluster in at one mode: Ohio Newspaper Poll (Obama +3), CNN/Time (Obama +4), FOX News/Rasmussen (McCain +2) and NBC/Mason-Dixon (McCain +1), which averages out to Obama +1 – a clear tie.
The other 4 cluster at another mode: Politico/InAdv (Obama +10), Big10 Battleground (+12), Quinnipiac (+14) and Suffolk (+9), which averages out to Obama +11 – a romp.
What this tells me is there are two classes of turnout models in play. Â One is the traditional model using historic turnout models to weight their results. Â They reflect the Gallup “traditional’ model which shows a very tight race. The second set of polls is using the untested or unverified turnout model which tries to claim that willingness to answer a poll has a strong and equal relationship across party affiliations (and other factors) with willingness to vote. These polls are analogous to the Gallup ‘extended’ model – which could be (and I am almost willing to predict is) pure fantasy.
Are McCain and Obama tied in Ohio in a close race, or should Obama be measuring the drapes in the oval office? For McCain-Palin supporters I can tell you the only way for the fantasy polls to be right is if you folks give up and sit home or waste your vote on a 3rd party candidate. Palin could put this away.
For example, look at VA as well, another state some people claim is gone for McCain (even though Palin will hit 3 big areas in the state Monday). Again we have two polls which are clustering around a close race: NBC/Mason-Dixon (+2) and SurveyUSA (+6) for an average of +4 Obama. Two others show a Obama blow out: CNN/Time (+10) and Rasmussen (+10), resulting in a +10% Obama lead. We can find the same pattern showing up in state after state like Florida, where one poll mode rests at a dead heat with McCain +1 and another is sitting around Obama +6.
RCP simply averages these polls, but I think the reality is one mode or another is going to be the right answer. Nationally it is either an Obama +4% close race or an Obama +10% landslide today. It is not in between these two modes. Unlike normal bimodal statistical results, where there are really two outcomes likely, in the political polling world only one mode will proven to be right, and the other will have been proven to be wrong. Which one will survive is actually up to the voters now.  But the close race mode would mean it is very easy for McCain-Palin to pull off a shocking win on election day. And as I noted early, early voting poll results from Gallup tend to show this is a close race, not a walk in the park for Obama.
Any idea what the 2004 early voter totals in Florida were for Bush/Kerry?
AJ,
I like your site and you seem like a nice guy. And I know what you’re trying to do with the polls. But, I think you should stop poring over them. The polls are almost certainly wacky, but they range from McCain being a little down to McCain getting wiped out. It’s not looking good.
That said, both what Obama is and what he represents in the larger scheme of things are frightening on so many levels. He’s all wrong for the U.S. We just have to keep fighting right up until election day.
The odds are stacked against us, but I for one take some satisfaction in the fact that, regardless of how this turns out, I will have understood fully who this man is. Furthermore, I will have tried to warn people (some people in my own family are voting for him) what is happening.
I think you need to just forget about the polls. They don’t matter at this point. It’s not about winning or losing at this point; it’s just about fighting for what’s right. That’s why I have contempt for the backbiters in the Republican Party and conservative movement who are jumping ship.
Just keep slugging it out, come what may.
Danking,
+5 Bush
Cochino,
I am not really that much into the polls as explaining what they mean.
And they are showing bimodal results, which is not possible. Either Obama is heading for a landslide or a tough day. The data is what it is, I cannot change it by simply looking at it and explaining what is says.
cochino,
you are correct just go vote. By the way, any person who doesn’t vote because of low poll numbers is a fool. By the way I also feel nothing but contempt for people who sacrifice their beliefs merely to curry favor.
robert:
Me too.
No one seems to be factoring in “likely voter” behavior in low turnout states like New York to Electoral College results.
Or whether the Dem machines are working on higher turnout – tidbit buried in today’s news that Philadelphia Dem machine has been totally ignored by Obama.
No one seems to be factoring in “likely voter†behavior in low turnout states like New York to Electoral College results.
Hmm.
Wonder how many PUMAs there are in New York?
Or whether the Dem machines are working on higher turnout – tidbit buried in today’s news that Philadelphia Dem machine has been totally ignored by Obama.
YGBFSM!
verdi and cochino, right on!
I live in La, it is supposed to be a blowout for McCain, and I’m voting for McCain. I would vote for him even if it were supposed to be a blowout for the marxist. vote early and vote often, as the Dems say.
Also, I think the polls are using traditional data which says a certain percentage of registered Dems are going to vote and vote Dem. I think the PUMA’s are gonna demolish that model.
Yes, I wonder what the PUMA’s and the Chaotic Ones are doing to the poll numbers….
We shall see.
Did anyone get to participate with Jerri Thompson on the TeamSarah call today? 1 million women were to be phoned today.
Go Sarah! Go John! YOU CAN DO IT!
Listen, guys. Let’s be real. We’re probably going to lose. Just go out and vote! If we don’t do that, we’re sure to lose. Much of the polling (the stuff that shows Obama up by double digits) is ridiculous. Nobody believes it. It’s thrown out there to make you give up.
I live near DC in Maryland, a state where Obama is going to win big, but I’m going out and voting for McCain. As I said in the post above, I’m working on some Obama supporters in my family. They live in Pennsylvania. I figure, if I can flip one of them, I’ve done my part.
I’m not going to look back 30 years from now and tell my grandkids that I didn’t bother voting in the election that sparked a socialist wave that washed over the country because ‘Obama was so far out in front that it didn’t seem to matter’. No way. My vote is going in the “against†pile. If that pile’s too small to carry the day, so be it. I’ll be able to hold my head high.
Maybe this is how it felt in 1980 in Reagan’s first election, the sense that this year is a big turning point. I was too young to vote in 1980, but it was in the early 80’s that I began taking an interest in politics. I’ve never felt this way about an election before. I have the feeling the country is at a crossroads. And the choices are stark.
Just keep fighting right up to election day. Screw the polls.
Cobalt Shiva: it is not necessarily the PUMA’s that will determine New York outcome. It is the socio-demographics of the voters who actually do vote in a mostly low-turnout base. In 2004, the 13 CDs in NYC had average 125-150,000 voters, less than half the voters of a competitive CD.
Only one NYC CD is competitive in 2008 – the sole Republican CD that Fosella gave up. Way upstate New York, New Jersey and Connecticut each have several very competitive CDs. I just read the NYT’s lackluster endorsement of Lautenberg for Senate (NYT did endorse one Republican each in NJ and CT for House), so who knows what will happen in the tri-state area where many voters actually do care about capital gains tax rates. All three states skew older and more affluent than average.
The NY Dem machine relies on low turnout to protect incumbents, and Hillary has little sway with the machine. If the Obama campaign has no NY (or CT) GOTV organization, then how will the reliables actually vote? Start with 1.5 million Jewish actual voters (and a whole lot more Catholics) in New York…
Massachusetts is an entirely different socio-demographic dynamic (49% of voters are unaffiliated with a party), but also will depend on turnout. Gov Patrick’s campaign by Axelrod was the test run for Obama’s hope and change message, and Patrick may be climbing back up to 45% approval after a disappointing start. There is a ballot initiative to eliminate the state income tax that should drive turnout.
I could go on, but my point is that the different poll models MAY need to incorporate a hybrid between the traditional voter model and anticipated voter model.
So much effort in the usual swing states and VA + so little effort in the usual safe states may = a surprise for everyone.
BIRDALONE,
AS A NEW YORKER I AM VERY SKEPTICAL OF THE NY GOP.
COCHINO,
RIGHT ON.
I know it is Zogby, but they have a poll out 49-46 with Obama on top. The IBD poll had within a point or two.
I think it is a long shot, but McCain can still win. I hope. Whatever, I am going to vote.
Cochino,
Great post.
Already voted McCainPalin in Blue California. Never cared about the Polls but how can you ignore them when you are bombarded with the numbers 24/7. Even the Yahoo browser page, every day for the last month there is a negative McCain or Palin story or something glowing about Mr. Obama.
I feel like I am in a 3rd world country.
Tom
PS I believe I just changed my sister-in-laws vote in Indiana after explaining just how the tax policies will effect her pocketbook.
Terrye
Zogby does have the race closer. Just vote and see.
I do not have a background in polling statistics, but I do have a background and a familiarity with biostatistics. Both deal with people in population. Here are my thoughts regarding the RCP poll average. I think it is erroneous to be averaging polls using the polls themselves as a unique data point each. What the RCP really should do is a true meta-analysis and increasing the data points. Each poll should be weighted based on the numbers of people sampled. Then calculate the their polled choice (McCain, Obama, & Undecided), their assignment of the political parties, and their margin of error as weighted by poll size. A poll that has 600 people should not be simply averaged with that of one that polled 1200 people. A poll that weigh 35% democrats cannot be simply averaged with one that weigh 40% democrats. Finally, the RCP should not be expressed as an average difference as this is meaningless without considering the error margins. I think it would be better to report a statistical tie as a tie rather than a raw number without statistical references.
[…] mine. Â The turnout model scam of 2008 will be the big story of this election. In my last post on polls I noted how many of the polls out there are starting to fall into two camps. The first camp is the […]
[…] may be way off based on optimistic voter turnout models which are historically nonsense (see posts here and here). One of the first real indicators of whether Obama really is enjoying some massive lead […]
[…] may be way off based on optimistic voter turnout models which are historically nonsense (see posts here and here). One of the first indicators of whether Obama really is enjoying some massive lead is the […]