Bush's free falling poll numbers (graph)

[Czolgosz]

Thanks, as always, to pollkatz. If you like the graph, please kick it high.

[iconoclastNYC]

1. They'll still make him into a saint?

[CoffeeCat]

2. I like it...

..It's kinda Mondrian--ish...

:)

[Czolgosz]

20. Here's your very own Mondrian generator (fourth amusement from the left)

<http://www.andyfoulds.co.uk/flash_design.html>

[CoffeeCat]

30. Kewwwwl...that's like therapy. Neat site and great work...

...thanks for sharing that.

[Czolgosz]

31. It also has a feature that translates any word into Bushspeak (it's the

fourth amusement from the right).

[lpbk2713]

3. Looks like a ski run from the recent Winter Olympics.

:rofl: :rofl:

[rpannier]

4. Excellent Graph

Cuts right to the heart. I'm sure when scrub sees it, his flunkies will tell him it's a disapproval rating chart.

[usregimechange]

5. Wow! See any outliers?

[Lerkfish]

8. abc and fox consistently rank him too high, it seems.

[Czolgosz]

13. Gallup also consistently inflates *'s numbers

[wtmusic]

6. His three little spikes

1) The US gets attacked
2) He starts an unnecessary war
3) He catches Saddam Hussein while costing 100,000 Iraqis their lives

What a hero.

[Hissyspit]

10. They are all artificial.

Fictitious President. Everything about him is fake.

[TomInTib]

11. That's all they have. n/t

[Czolgosz]

15. Makes me think Cheney and Bush are debating whether to start another war

or to allow another attack, because they sure as hell aren't close to catching bin Laden and they could use a bump in the polls just about now.

[funflower]

7. We need a bunch of independent pollsters to do exit polling in November.

It would be interesting to compare this kind of "spread" with the official results. :beer:

[aquart]

9. My favorite confetti.

[Breeze54]

12. Set to music, it would snd like a durge...(sp?) er...funeral march! n/t

[linazelle]

14. Gee Zogby's at 40--not a surprise, Zogby is very biased. nt

[tuvor]

16. It's fun to extrapolate. Looks like he'll be at 30% in October. n/t

[Czolgosz]

17. I like the look of his coattails in November...

[question everything]

18. Excellent. A visual that even a literate-challenged like Bush

can understand

[intheflow]

19. Looks like Mardi Gras confetti raining down on *'s parade.

Wah, Georgie! :rofl:

[Canuckistanian]

21. What was that Harris poll at the beginning of the year?

It looks like it's at almost exactly the same level as the CBS poll.

[Czolgosz]

22. It was also a 34%

[AndyTiedye]

23. At This Rate of Decline, By 2008, His Approval Rating Will Be About -5%

They are going to need a lot more of these:

[Clarkie1]

24. Looks like a secular bear market for *

with cyclical "bulls" (a very apt term in this case)

In other words, the overall trend is downward, with a cyclical (bull) uptick now and then...

[Czolgosz]

52. It also looks a lot like the inverse of the price of gas:

Thanks, once again, to pollkatz.

[Clarkie1]

61. Facinating. n/t

[Czolgosz]

25. Here's Bush's numbers compared to Nixon's and contrasted with Clinton's

Again, thanks to pollkatz.

[Clarkie1]

27. Interesting how Clinton's numbers rose in second term and were stable. n/t

[Nikki Stone 1]

26. But these polls come from "the reality based community"

These guys don't care about reality.

Or maybe they know that despite disapproval, the public won't or can't do anything.

[Czolgosz]

32. BushInc may not care, but the 'Pukes running in '06 are plenty worried!

[Clarkie1]

28. e-mailed to white house: "Happy Mardi Gras..how do you like the confetti?"

:-)

[Czolgosz]

29. lol

[ProfessorGAC]

33. Folks, I Got These Numbers In Raw Form

The fit of the data v. time is unbelievably correlated. The only two upticks are based upon 9/11 and the invasion of Iraq. Yet, if you discount those, you will find that the negative slope indicates he would be at about 33% approval, overall. The r^2 for this fit is almost 92%

IOW, if you ignore the 3 month periods following those events, and just use that data to plot approval ratings v. time, you find him at 33%.

With those upticks, he's at 34% approval. The r^2 for that is around 88%.

So, those upticks were strictly americans being americans, and DID NOT HAVE ANYTHING TO DO WITH WHOM WAS IN THE OVAL OFFICE! Anybody in office when 9/11 happened would have had that huge bounce, and anybody in office is going to have an uptick when a war kicks off.

He was NEVER a popular president! And, his popularity has decreased with time along an EXTREMELY predictable curve.

The Professor

[Czolgosz]

35. Great analysis. Where'd you get the raw numbers?

[ProfessorGAC]

36. Through A Network Of Statisticians And Analysts

I know people at all the major polling companies who work as analysts and stat people. Some of them, i taught. Others, i know through organizations.

I can get pretty much any data i want, except of course, stuff the gov't lists as secret. If it's public, even if it's the intellectual property of one firm or another, i can still get it. I just can't use it to make money. But, fair use allows me to model data from any source for academic or publishing purposes.
The Professor

[underpants]

41. There is also a small uptick around Saddam's capture and the '04 election

I just wanted to point that out.

I have said it before and I will say it again now--the best thing for W is when lots of Americans die. Sad to say but true.

Nice how little of an uptick the election had :think:

It looks like a Christmas Tree laying on its side. If you see the DISAPPROVAL numbers in the same graph it looks just like an XMAS tree.

[ProfessorGAC]

43. Neither Of Those, However,. . .

. . .are statistically significant. Neither tick is more than 0.2 sigma from the overall mean of the time weighted moving average. With this averaging, the standard deviation of the dataset is pretty small, and at 0.2 sigma (+ or -), the probability of any single weekly value moving that far from the mean is about 96%. So, within 96% chance of being right, we can predict any value within 0.2 sigma.

Now, as time goes on, the dataset gets larger and as k goes up, and degrees of freedom increase, the variability of any single point shrinks. So, this is an even smaller margin of error today than in those two periods.
The Professor

[underpants]

49. Well yeah of course that goes without saying

You hurt my brain

[Beetwasher]

55. LOL!

Is that hard to do? ;-)

[happydreams]

57. With an approval rating so low how did he "win" the election?

:eyes:

[Mugsy]

34. Trend puts him at zero around February 2008. Below 20% by October.

Should make the elections interesting.

[ProfessorGAC]

38. Not Quite!

I think you're kidding, but the current trend would put him at 25% in November. However, since there is a segment of the citizenry that is addicted to the kool-aid, the numbers will start to bottom out, asymptotically toward a value that reflects the nutjobs who won't turn their back on him no matter how many more things he screws up. Near as i can figure, that number is around 22%. So, he'll never actually fall below that, and the rate of change will becoming increasingly slow as it approaches that.
The Professor

[Czolgosz]

39. I wonder of the "bottom out" number for a Republican is about 37% and Bush

dips below that number when he does his damnedest to piss off Republicans as well as all correct-thinking people (I see his deepest dips are when he annoyed Democrats as well as his knuckle-dragging base with the Miers nomination and the UAE port security debacle).

[ProfessorGAC]

40. As Near As I Can Figure. . .

. . .and it's from lots of data for decades, smoothed out to work with weighted averages, it looks like there is about 60% of the republican base that would be fairly termed radical right.

As a result, these are the folks who won't abandon him, no matter what, because the alternatives are all "EVIL". As a result, 60% of the 3/8ths of the people that consider themselves republican or repub leaning independent (no matter their actual registered affiliation) will stick with him.

60% of 37.5%, is 22.5%. So, that's where i get the 22%. There are some assumptions, but they are supportable through the data that drove the conclusions.
The Professor

[Czolgosz]

42. I think we agree that the GOP base is about 3/8 of the voters so whenever

Bush slips below 37% you know he's wading into Nixon territory.

[ProfessorGAC]

44. He's Toast

He is now a true lame duck. No agenda, no support, repubs in congress putting space between themselves and him, and the people no longer supporting him. Congress could take some extra vacation time.
The Professor

[Leopolds Ghost]

47. How exactly does it matter, now?

No way to impeach him even if Dems got a narrow (DINO padded) majority in November.

Cheney resigns just allows him to appoint an heir apparent.

Americans have short memories and are endlessly forgiving of celebrity.

[Czolgosz]

51. It matters because Bush's unpopularity will be an anchor on GOP candidates

[Beetwasher]

56. Cheney's At 18%

Which would lead me to believe that that may be the ultimate bottom.

[Czolgosz]

60. That's a good point. Why aren't the rabid dogs showing Cheney more love?

[Clarkie1]

62. self-delete.

[alcibiades_mystery]

37. It will almost be an entire year that he is below 50% approval

When will the lapdog media stop calling him a popular President?

How long will it take for more to disapprove of him than approve of him for anyone to mention it???

[Neil Lisst]

45. that charts says war with Iran is the next spike for Bush

we can hope it's wrong

[Czolgosz]

50. Shhhh. Not so loud. They're listening.

[Digit]

54. I doubt it would work again....Americans are war weary.

...not to mention how much further in debt it would take us.

[Leopolds Ghost]

46. Anyone else think the incredible tightening of distribution pre & post-911

looks incredibly suspicious?

Distribution of the polls pre-9-11: All over the place, within 15% of one another.

Distribution of the polls ever since 911: Incredibly tight, usually 7 points separating the highest from the lowest poll.

And the fact they are in lockstep suggests to me that "new polling factors" or new means of asking questions, or general fear of giving incorrect answers was introduced after 9-11. Very ominous in a supposedly free society.

[mdmc]

48. very nice

thanks for the data:kick:

[paineinthearse]

53. Look at that spike July-October

That's when KKKarl was in full charge, things really have slipped since he has been paying attention to his own (doomed) future!

[lonestarnot]

58. Not the bushitler slimey slope again...

:toast:

[npincus]

59. looks like confetti on his impeachment parade