Gallup: 49 Romney 48 Obama; they call it a "tie"

[VirginiaTarheel]

[jezebel]

1. where are you seeing this? thanks.

[Enrique]

2. margin of error n/t

[yellowcanine]

3. Only a "tie" if it is a one time poll.

Not a tie if it is a tracking poll or when one considers that there are several other national polls showing the same trend.

[geek tragedy]

11. Gallup is showing Romney ahead, which conflicts with every other non-tracking poll. nt

[Mass]

4. Did they give a result or are you guessing?\nt

[MadBadger]

7. its coming from Editor Frank Newport

[DemsUnited]

5. Way better than last R+5 tracking poll which RCP is STILL using

national momentum in the last week still going to our Prez even in outlier Gallup

On edit: RCP is still using the old Gallup R+5 tracking number in their aggregate. if they ever decide to replace it with this new R+1 number, Obama's lead should go UP.

[BeyondGeography]

6. They had to come down off their cliff

Gallup and Ras can both say they were well within the MOE when Obama wins.

[Mass]

8. It is a tie, given the MOE. And it is one of only 2 national polls giving Romney ahead.

[gcomeau]

17. No it isn't.

It's nearly a tie, and Gallup has been out right ever since they introduced their insane LV model after the debates so it doesn't concern me they have Romney up by only 1 at this point... but it is NOT a tie. A tie being within the MOE does NOT mean the poll says the race is tied. I am so sick of people saying that. It's the same thing the press uses to portray the race as far closer than it is when Obama has been leading it clearly all year long.

A tie being within the MOE means it is possible, somewhere within the 95% confidence interval, that it MIGHT be a tie. But if a poll says one person is up 1 the poll says that person is up 1. The end. Not tied.

I hate it when they do that the other way on the majority of polls saying Obama is winning, let's not do it in the other direction ourselves. Encouraging proper treatment of polling data is the only way we might, maybe, eventually move past media horse race manufacturing.

[Mass]

19. Sorry, I call it as I see it, This poll shows Romney by 1, which, given polls are not exact science,

(you cannot account for people who answer the polls lying, for example) means it is too close to call. I understand some Romney supporters can feel hurt when people say that, but I am so sick of people assuming each independent poll is meaningfull. There are not.

[gcomeau]

22. Go ahead and say it's "too close to call" then.

But saying the poll says the race is tied is factually, objectively, wrong. It's math. 2+2 =/=5. Gallup says Romney is up by 1, not that the race is tied. PPP says Obama is up by 2, not that the racer is tied. Etc...

The polls say what the polls say. you can make arguments that they're wrong,. And in Gallup's case the arguments are pretty good. But that's Gallup being wrong, not Gallup SAYING the race is tied. Ok?

[Mass]

24. Complain to Gallup. They called their poll a tie.

But sorry to have upset you.

[gcomeau]

25. Of course they did. It's the popular thing to do and it covers their asses.

That way if Obama wins they say "well, we had it as a tie so who knew who was going to win? Guess it was Obama, doesn't reflect poorly on our polling! No sir..."

It's still wrong. I blame them for saying it. I blame the media for saying it. I blame *anyone who says it* for saying it, and I point it out, so hopefully they stop saying it and it stops spreading and maybe one day in the distant future we can get factual, accurate election reporting.

[blueclown]

27. Are you aware of how a sample works?

I don't think you understand how sampling works.

[gcomeau]

28. Yes.

I'm also aware of how Gaussian distributions work.

If a poll says a candidate is up 1, it says they're up 1. A tie falling with in the 95% confidence interval of the distribution does not alter that fact.

[VirginiaTarheel]

9. Gallup and Ras, the two most R-friendly pollsters, can only manage a +1 the day before election day.

[TexasCPA]

10. Gallup sucks

One less talking point for Faux News.

[VirginiaTarheel]

12. So Gallup says it has gone from Romney up by 7 about a week ago to only up by 1?

So they are claiming there has been a six or five point swing in a week?

[geek tragedy]

13. They're still going over the waterfall with (R)-Assmussen.

The only two pollsters who will have picked the popular vote loser.

[aaaaaa5a]

15. Didn't you read your right wing talking points?

Romney was up 7 until Sandy hit.

[Jennicut]

21. Yup. A 7 point swing in a week. They had to do it or be seen as totally outside the norm of the

other final national polls. So misleading for them this election as they have been totally outside the norm for weeks.

[zebe83]

14. I don't see this poll mentioned on Gallup site as of yet...

And Romney was +5 with them on their last poll on 10/28.

[bluesoul]

16. new Gallup poll

It will be, but its more or less official:
http://livewire.talkingpointsmemo.com/entry/gallup-national-poll-romney-49-obama-48

[zebe83]

18. Thanks....

That will make the RCP average go up for PBO with the +4 swing.

[progressivebydesign]

20. Well considering we DON"T use the popular vote, who really cares?? nt

[libguyob_19556dd]

23. Gallup is a joke

Period

[hrmjustin]

26. I agree and welcome to DU!

[VirginiaTarheel]

29. Full Gallup results-Likely Voters-R49%(-2) O48%(+2) Registered voters-O49%(+1) R46%(-2)

[TexasCPA]

30. You got to love RCP

Since they had to pull the Romney +5 Gallup number off, they made sure to the the National Journal poll that had Obama up 5 off their list too. RCP is such a joke.