Cruz at 31% Rand 3% DM Register poll

[groverblue]

I've gone through how horrible the polling is

http://www.ronpaulforums.com/showth...etime-Republican-Debate&p=6065620#post6065620

Near the low (or the high) end of the polling range, the error bars get a little funky. So if you're at 4%, instead of having a standard error of +/- 2%, you will have lopsided error bars. So for example, +2% -1%. This is because the "standard error" is based on the width of a gaussian distribution. This gaussian is an approximation of a probability distribution near the middle of the range of a large data set. If you get near the edge of that range, the gaussian "bell curve" is no longer a good approximation, instead it's better to use something known as a "Poisson distribution".

In this image, you can see the Poissan distribution (lamda = 1) slowly morphing into the Gaussian bell curve (lamda = 9) as the median of the curve moves away from the edge of the range.

A poisson distribution is lopsided, and therefore should have nonsymmetrical error bars around the median

I didn't understand any of that. haha. Please explain.

 

[65fastback2+2]

I didn't understand any of that. haha. Please explain.

margin of errors only work in gaussian curves where the curve is normalized (equal on each side).

In other words, a margin of error of 5% for someone polling 30% is "acceptable" but for someone polling 5%, is unacceptable.

 

[Joeinmo]

Easy to beat Cruz on blogs, let people know that CRUZ is a Bushie, worked for GW Bush, wrote a 15 pg love letter about him.

http://www.gq.com/story/the-lost-love-letter-ted-cruz-wishes-he-never-wrote

its about replacing Trump with Cruz, it's what the establishment wants

 

[BUSHLIED]

I doubt Cruz is truly leading in IA. The poll simply reflects the establishments attempt to marginalize TRUMP whom they also detest. Political tactic like they did with Ron. If you look at the polls for Ron in 2008 and 2012 he exceeded those numbers in the final vote tallies.

Ron in 2008 in IA polling at 3.5% finished with 9.9%
in 2012 in IA polling at 11% finished at 21.4%

I suggest the true for Rand. Also you'll notice that the front-runners under performed as both Cain and Romney did not come in first. in fact, look at Santorum winning while he was polling at 3.5.

Polling is clearly subject to error and bias.