GALLUP: Generic poll had faceless GOP beating Obama 47-39%. In August, it's Obama winning 45-39%

[jefferson_dem]

:bounce:

http://www.gallup.com/poll/election.aspx

h/t: http://twitter.com/#!/EricBoehlert

[Broderick]

1. And people are running around here

actually thinking Obama will lose!

Good GRIEF folks!

If I were a gambling man, I would quote a Seinfeld episode for myself. IT's A LOCK JERRY!

[FrenchieCat]

2. Finally!

We're getting somewhere! :thumbsup:


Read and share.....http://joewo.com/WordPress/?p=9293

Future, here we come! :bounce:

[jenmito]

3. K&R! The "Generic Republican" is his strongest competition. n/t

[Tarheel_Dem]

4. May I join you? Especially since this was much cause for "concern" last month.

;)

How long before his detractors chime in to tell us, we're over a year out from the election? Cause when the numbers were reversed, apparently the sky was falling. :rofl:

[tularetom]

5. IMO the republicans punted on the 08 election

they knew that whoever won would inherit all the fuckups of the cheney/bush administration (and get the blame for them) so they put up a weak ticket and sat back waiting to snipe at the Democratic president for failing to turn things around. In the process they scared off any capable and/or sane potential candidate who may have opposed Obama, leaving the field to the insane clown posse that now makes up the crowd of possible republican nominees.

On balance, Obama has been a bit of a disappointement but he is so superior to this group of loonies he will be elected with ease.

[PollyPeptide555]

7. Insane Clown Posse? Lol!

[joshcryer]

6. With the exception of outliers and whatnot, he'll gain a point a month.

The question remains how the Congress will fare.

[CakeGrrl]

8. But, but, some of DU says that the ONLY way Obama can win is if

Palin runs. That's the ONLY one he'll be able to beat.

Of course, those are the more generous assessments than "I can't possibly see how he can win in 2012."


:rofl:

[Davis_X_Machina]

9. He'll be doomed right up to his inauguration.

[jenmito]

10. Largely due to the

Public Option thing, I'm sure.

[Davis_X_Machina]

11. That, and the Rick Warren thing..

...and Van Jones. And Elizabeth Warren.

I've rarely seen a man so doomed.

[Number23]

16. lol!

:thumbsup:

[LiberalFighter]

12. A generic poll would beat the Gallup poll any and every time.

[ThoughtCriminal]

13. They've seen the available faces

Like 2008, it will look close or even a Republican lead after the GOP convention, and then the spread will grow in favor of president Obama after a short bump. They are not going to be able to back-peddle the crazy.

[former9thward]

14. In July the poll was denounced as being unreliable and worthless.

Now it is a good poll and accurate. We will see more of these flip flops as the election gets closer.

[FrenchieCat]

23. So who are you rooting for?

I'm sure at FR they will call this poll unreliable and worthless,
and the July poll solid gold.

And? :shrug:

[totodeinhere]

26. Are you saying that we should only believe polls that show President Obama winning?

And we should disbelieve any polls that show him losing? To me that attitude would be intellectually dishonest. And the possibility that freepers might do the same thing in reverse doesn't make it any better.

[FrenchieCat]

34. the only reason I care about polls is because they tend to have undue influence....

Which is exactly why media outlets pay for such.

The question and the data is easily manipulated, but for some,
it appears so "official" as though a disinterested 3rd party
is doing their due diligence. The MO is only mentioned if it helps
whatever the media wants to convey. Also, folks don't realize that polling
companies can call a particular zip code at a particular time of the day
and get exactly the polling results they want.

I have noticed that it isn't only till about a week prior to
election day that polls start to be accurate, and even then there's always
the outlier to be used to manipulate voters some more.

I remember during 2008 election, the Pennsylvania poll that showed McCain ahead
or tied or some shit; and then Obama beat him by what? 12 or 16 points or something
like that?

But most of all, polls as a tool for pundits is a very powerful thing to use on voters
who like to be in with the cool crowd, doing what everyone else is doing. And there are many of those types of voters, and the worlds largest clients of polling companies understand this very well.

[former9thward]

39. I don't "root" based on polls that are over a year before the campaign

really starts. They are nonsense and anyone who puts stock in them will be mistaken. My only point is the habit on DU of attacking polls when the results are bad and saluting them when they are good.

[DCBob]

15. long way to go but this is encouraging considering the mess this country is in.

Bottom line.. as long as we dont fall back into another recession Obama wins.

[dsc]

17. This represents an improvement that is within the MOE

so hardly time to celebrate. Now, that said, I will say what I said last month. These numbers are troubling. Combined with the state of the economy they are cause for worry. By no means is this a fate accompli but he is in trouble. Currently he has the benefit of, with the exception of Romney and Huntsman, a bunch of scary and radical opponents. If the Unemployment rate falls to around 8, then he is likely a shoe in. If it falls to 8.5, I think he is likely OK. If it doesn't fall, I am honestly more than a little scared. He could win by his opponent being just too awful to bear but that is a pretty rare occurance. I can only think of a few elections where that happened. One is Major's win in Britain. Arguably one was Truman's in 48. Carter thought he had one in 80 and we know how that turned out.

[jenmito]

19. Ummm, no it doesn't. It's a complete reversal. The first had the GOP candidate ahead and

the later poll had Obama ahead.

[dsc]

21. The margin of error is 4%,and works either way

therefore a 39 in a poll can be as high as 43. Similarly a 45 can be as low as 41. His improvement is at best barely above the MOE and is, I believe, within it. On edit I can't find the MOE. I am not sure if we have to pay to have access to it or what the deal is. 4% is fairly standard. If the MOE is 4.5 then this is within the MOE. If it is at 4% it is just barely outside. I will concede it is more likely than not that he made a slight amount of gain from July to Aug but the notion we can be sure he has reversed the lead is just not something we can say. One last edit I found the MOE for approval numbers, and that is 3. If they have the same number for that poll, not an unreasonable assumption, then they are likely outside the MOE. GOP was 47, could have been as low as 44. Is now 39 could be as high as 42. Thus it is possible, not likely but possible that the GOP only fell by 2 points. Similary Obama was 39 and could have been 42 and is now 45 and could be as low as 42. Thus he may not have approved at all on his side.

[jenmito]

25. "Generic poll had faceless GOP beating Obama 47-39%. In August, it's Obama winning 45-39%."

Obama rose 6% and the generic Repub. fell 8%-a 14% swing-HARDLY within the margin of error.

[dsc]

27. Obama's 6 is within the MOE and the GOP's 8 is barely outside it

It isn't my fault you don't know how MOE works. The fact is that the MOE is a range and thus is a range of 6, not 3. I was wrong on what the MOE was, but had I been right the GOP loss would have been within it. Again, I am not making shit up. I teach this for a living. I know this kind of stuff. The fact is he is outside the MOE but not by much.

[jenmito]

30. It's a 14-point swing. It's outside the MOE. The 3+-MOE means it could be 3 points higher or 3

points lower than the number. He went 6 points higher AND the Repubs. went 8 points lower. Try to justify your mistake all you want to minimize this big change. It's not working. Even if you now say it WAS outside the MOE but "not by much." :eyes:

[dsc]

33. You are just plain, flat out wrong

A 3% MOE means that EACH number can be as much as 3% higher and 3% lower. That is the way it works. It isn't my theory it is the way statisticians have done this since the beginning of statistics. It isn't some anti Obama plot. Thus, as I said in my previous post. Obama has not shown any statistically significant improvement while the GOP did show some statistically significant falling off. The fact is the poll in July was a statistical tie (right at the MOE) while this poll is a narrow Obama lead. He improved, but barely. Again, this isn't something I am making up. It is basic, stats 101. I am sorry you don't know it. It isn't my fault.

[jefferson_dem]

35. Sorry, that's not correct.

It is not correct to use a MOE X 2 formula to determine "inside or outside" when a variable consists of more than two value categories. In this case, since there is a segment of the sample (16%) that is "undecided" or supports "other", we must employ a formula appropriate for that multinomial distribution. The result would be more narrow than simply tacking on +/- 3 points for Obama and the generic GOP.

It is confusing, I know. I teach methods also but wasn't exactly sure so I looked it up. Here's a primer: http://abcnews.go.com/images/PollingUnit/MOEFranklin.pdf

Bottom line: the difference between these two survey results is easily beyond the MOE.

[jenmito]

36. Thank you! n/t

[jefferson_dem]

37. No problem.

Though I must say that the MoE argument just reeks of a pointless red herring anyway. :hi:

[karynnj]

38. Another problem is that there is a negative correlation to the two variables

Even if there were just two categories, they would not be independent variables. This wiki page has the math - scroll down to find the variance when they are correlated - looking at the formula you see the variance is smaller if they are negatively correlated (ie when Obama goes up, GOP goes down) because "rho" will be negative.

http://en.wikipedia.org/wiki/Sum_of_normally_distributed_random_variables

[totodeinhere]

24. And the next poll might flip again for all we know. Look for wild swings in the polls to continue.

The same Gallup poll on 8/17 had Obama within the margin of error against Perry. Romney, and Paul.

http://www.pollingreport.com/wh12gen.htm

[jenmito]

28. So? That has nothing to do with the fact that the OP's poll changes are NOT within the MOE. n/t

[totodeinhere]

31. Given the tendency of polls to fluctuate so much and given how early it is it's not really

that significant whether it's in the MOE or not.

[jenmito]

32. Not the point of my post. n/t

[jefferson_dem]

22. Nevermind.

Addressed above.

[Hippo_Tron]

18. Polls are fickle as a fuck, IMO this is going to be a very close election

And I hope everyone gets out to vote so that it's close, but in the right direction....

[totodeinhere]

20. This just goes to show how fickle polls taken now can be. That's why we really

shouldn't pay attention to early polls like this. The only poll that will really matter is the one taken on election day 2012.

[Mr Deltoid]

29. Perhaps the repubs should make a mask from a paper plate with eye holes cut out

And run somebody wearing it. Mr/Mrs faceless seems to be the front runner.

[progressoid]

40. Yay, another definitive poll!!11!

:eyes:

[jefferson_dem]

41. Not really.

Just a fleeting snapshot. :hi: