Obama Is Handily Winning...a little deduction from the polling

[kristyt]

Too many people use inductive reasoning, let's use deductive reasoning to figure this out.

Just as a refresher, deduction moves from general to specifics, whereas induction looks at particulars and tries to make a generality.

So lets start with assumptions (these are what you'll need to knock down)
1. Polling hasn't been reliable this cycle.
2. Polling companies want to get make correct predictions.
3 Given 1 and 2, polling organizations will be more cautious/conservative in their polling.
4. Getting the winner is far more important than the margin for political polling.
5. Obama has been in up over McCain in nearly every poll conducted since becoming the presumptive nominee (something like 44 out of 45 of the major organizations).
6. Obama is ahead by every analysis in electoral vote forecasting based on state polling...this is just what I've seen, perhaps you've seen others.
7. A wide variety of samples and weighting have been used, and the degree of Obama's lead varies widely poll to poll.

All this leads to the deduction that Obama is convincingly ahead of McCain.
Since this is deductive, the whole chain has to be valid, if the assumption is wrong at any point, the ultimate deduction is not logically sound.

If you wish to get more tenuous I think there are further deductions one can make.

8. Given the desire/incentive to pick the winner by these polling companies and the high volatility this year (at least in judging Obama races) they are probably constructing models that hedge against large margins either way, therefore making it appear to be numerically closer than it really is.

9. Since everyone measuring has Obama up, I think that's the crucial metric people should be using this polling for.

10. Finally, it's still early and things might change, but given the wide disapproval of Bush/GOP the high Dem turnouts in the primaries, the gap in enthusiasm for candidates in the parties, the betting sites have him up by large margins, the results of the special elections conducted this year, the great deal of attention paid to this race as compared to most Presidential races already....there is probably less chance for things to change as Obama is leading, and everything points to him having the advantages.


It's not over until the night of Nov 4th, but I think this race is likely already decided in Obama's favor.

What do you think?

[Diane R]

1. What do I think? I think we have to assume we're down by 10 points, and they are already cheating.

Enough claiming victory already. It's going to be an uphill battle to win against the cheating Rove machines, the spinning press and the skewed polls. This is no time to rest on laurels.

[Sugarcoated]

2. A good thing

at least in Ohio now there's a Democratic Governor and Secretary of State so it'll be harder to commit the kind of voter fraud and voter suppression they did in 2004. Florida, unfortunately, still going to be corrupt.

[kristyt]

11. Take it Which ever way Makes you most comfortable

Obama stresses the power of optimism and confidence. I don't think your Rx is necessarily for everyone.

There's a book that is pretty good called something like, "The Power of Pessimistic Thinking" (or something like that). The entire thesis is basically what you've described. Assume the worst and you won't be disappointed, you'll motivate yourself to work harder, etc. For many people this works.

But others do better with positive thinking. They are more productive, work harder etc.

So it's all about your own psychological makeup.

My post doesn't suggest you should take it one way or the other. Only that I think my deduction is reasonable that Obama is likely comfortably ahead in this race.

[MrModerate]

3. Another factor is that McCain is campaigning very badly . . .

And even if his folks get better at the job or they bring in more greybeards, they're not going to 1) get a lot better or 2) get past the fact that McCain himself is decrepit, declining, bereft of principles, and creepy.

What I want to see (and soon!) is Obama pulling away. We need about a 10-point advantage (a number I made up, but which seems reasonable to me), broadly distributed, to put this out of reach of the election fraudmeisters.

[silverojo]

4. Very true

In all the elections I've ever observed, the polls are damned near always correct. Every time people say, "Polls don't vote, people do", I roll my eyes, because that's delusional thinking.

The fact is that, so far, it is a close race. Hard to believe there are so many imbeciles who would even think of voting Republican after Bush, but it's a fact. Look at his ads during the Olympics--it's the same old "Higher Taxes" boogieman that he's invoking again...because it works.

Obama can't afford to miss a trick at this stage. We simply cannot have another Republican in the WH, poised to appoint rightard Supreme Court justices.

[Dawgs]

6. It doesn't matter what you've observed prior to this years Democratic primary.

The polls were all over the place. They weren't consistently correct.

And, the fact it, this is not a close race. An analysis of the numbers shows that winning by just 2% in the popular vote is a huge win in the electoral vote.

So, it's not close.

Everything else you say is correct.

[CTyankee]

7. Well, al lot depends on two things: how well represented each sample within the poll is and

how the poll is weighted. If the poll over-samples older people it skews to McCain, for instance. With regard to weighting, the 18-28 sample is weighted historically, but in reality you cannot accurately use data from 2004 in 2008, given the massive turnout of young voters in state primaries this past year. To say "young people don't vote" is ridiculous since we have data that proves that statement false in 2008. The idea that young people who voted in such large numbers in our recent primaries would not turn out for the General Election doesn't make much sense.

[newfie11]

5. How do the polls get these numbers

No one I know has been asked anything about the election.

[TexasObserver]

8. Very sound analysis.

If you take all the polls, toss the outliers, and look state by state, Obama wins with over 300 electoral votes. He holds hefty leads in many states. The national number of 43-44% that McCain typically gets is skewed upward by the hefty leads he holds in a dozen states. This means in the other 38 states, he has much, much less than 43-44%. He's not even close in states with 220 electoral votes.

McCain has to win almost every battleground state to win, and he won't do that.

We will see polls fudging their numbers and methods consistently to keep it closer than they really know it is. Media and pollsters have a vested interest in keeping media and pollsters interesting and newsworthy.

[Liberal_Stalwart71]

9. The important factors that are important to us are...

1. In *EVERY* so-called national poll, McSame has consistently polled no higher than 45%. In fact, he has hovered right around 42%-44% without much movement from there. That tells me that Obama has been moving much, much more and that's because the M$M has scrutinized him so much, under a microscope and has not done so for McSame.

2. These polls cannot account for new registrants, the power of the young voters, and the fact that no one really seems to pay attention until *after* the convention. That's why the Obama camp doesn't panic over these polls and neither should we. If after the convention, it doesn't move considerably or there's a sizable shift in either direction, then we ought to be concerned.

That's why it's baffling to me why many of us panic or become demoralized by the polls. That's why I cannot understand pundits like Rachel Maddow who is really turning me off with her whining about why Obama isn't doing this and why Obama isn't doing that.

We need to relax. Take a breath. Work hard and continue to register new voters, donate, canvass. But not panic each and every time these polls shift a little bit.

[kristyt]

10. Agree, But I Doubt We'd Ever See a 10 pt Avg Lead for Obama

We've already seen the LA Times and Newsweek both over that in mid June (before, it seems to my non-professional analysis, changed their sampling).

And I think those are good examples of what to expect. Even if a polling firm were to show a large lead nationally, because of the difficulties they've had predicting the winner in this year's Dem primaries, I think they'll like just adjust their sampling and weighting to make it closer.

Is this wrong of them? I don't know. On the one hand they have to admit the polls, where Obama has been concerned, have not been reliable. So, a large swing in Obama's favor is really sticking their neck out.

And, of course, the great example is New Hampshire. Where in a tiny state with massive amounts of polling by different companies, they almost all got it wrong (if not all of them) and by a substantial margin.

So my feeling is they'll all play it conservatively and we're not likely to see double digit margins.

[Liberal_Stalwart71]

12. That's o.k. because it keeps us on our toes and forces us to work harder!

Someone posted that the Obama team seem not to be bothered by the polls. Their internal polls during the primaries were spot on! They were only off by a few points in only a few states. So I think they know what the team is capable of. They have been absolutely flawless in terms of predicting their strengths and weaknesses. It'll be up to us to see it through.

[blondeatlast]

13. Until the last polls close on ED, we're ALWAYS down by 10 points.

Possiblyt the oldest election strategy known to man, but it works.

[kristyt]

15. That Motivates Some People

But not everyone. Some people won't bother if they think there candidate is behind, or won't be as enthusiastic.

Whatever works for you is good.

[AtomicKitten]

14. Obama has a stable and consistent lead.

I do believe he's got some smooth moves up his sleeve that we will see unfolding in due time.