Anyone ever read Real Clear Politics "averages" (VANITY)

2/11/16 | Self

[Riflema]

Countless News Organizations refer breathlessly to the Real Clear Politics "RCP Poll Average" as the go-to source for the big picture view of the candidates' standings. But did anyone take a look at that source and figure out how they come up with their numbers? With some trepidation, in fear of what I would find, I did.

Here's today's RCP Average and the individual polls it summarizes:

Poll	Date	Sample	Trump	Cruz	Rubio	Carson	Bush	Kasich	Christie	Fiorina	Spread
RCP Average	1/22 - 2/4	--	29.5	21	17.8	7.8	4.3	4	2.5	2.5	Trump +8.5
QuinnipiacQuinnipiac	2/2 - 2/4	507 RV	31	22	19	6	3	3	3	2	Trump +9
Rasmussen ReportsRasmussen	2/3 - 2/4	725 LV	31	20	21	5	4	6	3	3	Trump +10
PPP (D)PPP (D)	2/2 - 2/3	531 LV	25	21	21	11	5	5	3	3	Trump +4
IBD/TIPPIBD/TIPP	1/22 - 1/27	395 RV	31	21	10	9	5	2	1	2	Trump +10

Now take a look at any candidate's column and do their math, RCP style:

e.g., Trump: 31%, 31%, 25%, 31%, add 'em up and divide by 4 to get the mean, bingo!, 29.5%!!

I'll simplify the problem with this:- suppose one of those polls was a sample of 2000 voters and one was 100 voters and they came in as follows:

Poll	Date	Sample	Trump
RCP Average	1/22 - 2/4	--	22.5
Poll 1Quinnipiac	2/2 - 2/4	2000 RV	30
Poll 2Rasmussen	2/3 - 2/4	100 LV	15

Using RCP's technique they would give the Donald 22.5% (30% + 15% divided by 2).

But are you ready to buy that? What Real Clear Politics is actually demonstrating here is innumeracy.

If 30% of 2000 voters and 15% of 100 voters favor Trump, the correct average is:

(30% x 2000) + (15% x 100) = 600 + 15 = 615 voters out of 2100 = 29.3%

It's called a weighted average, RCP. The poll of 2000 voters has way more weight than the one of 100 voters. Get it?

I guess since the majority of consumers of this junk are journalists we should not be surprised that it is swallowed and regurgitated so readily. Over all my years of reading their junk, I long ago ceased to be amazed at their fundamental inability to master any of the hard sciences, you know, the ones that involve math. Sheesh.

As John Huang used to say, that's just my 2c. (and excuse the crude HTML!)

[Riflema]

Given the similar decline in editing in general, is it any wonder the Fourth Estate finds itself in such dire straits?

[LS]

Which may be why Trump under-polled in NH.

[scooby321]

Cruz is Push Polling in South Carolina

[Bidimus1]

Which would not affect RCP in any way.

[Durbin]

It’s called the RCP Average because they usually take 3 polls, and average them. That’s all. Not really sure what the big deal is.

[PIF]

Trumpeters ever holding on to the myth that DJT is some sort of wunderknabe, rather than the socialist tyrant in waiting he really is ... and I dislike all the other losers in both parties.

Very sad the country finds itself in the dire straits it has entered. Glad I’m as old as I am, seeing US politics come to the sad state it has fallen into by way of an electorate that is more interested in entertainment than civics, and now unable to tell anything about anyone unless they are told what to look for - usually by way of the politician’s own rhetoric and wishful thinking.

[Riflema]

Sorry, I think you missed my point. Their “average” is wrong. That’s the point.

[Anitius Severinus Boethius]

Yeah. It’s an average of polls.

If you want to do more in-depth statistical analysis of the polling data, feel free. But it’s a nice quick view at an average and a good place to find quick links to all of the major polling.

[outofsalt]

Double posting does not increase post accuracy.

[Durbin]

“Sorry, I think you missed my point. Their “average” is wrong. That’s the point.”

30+15=45
45/2=22.5

Looks like the average is right to me.

[bigbob]

Thanks for taking the time to lay it out. I’ve known it, just being the type whose motto is “In God we trust, everyone else bring facts and data and be prepared to defend it”. ;-)

To be honest, I’m not sure some polls are anything but smoke and mirrors BS intended to influence public opinion. All it takes is a fancy-sounding credible name to be in the polling business, right? So what’s to stop “Free Republic/University of Hardknocks Polling” from announcing the results of their latest polls without doing anything more than put the desired numbers into a press release? Nobody vets anything nowadays, and I’ll bet you it would get picked up and publicized by organizations who liked the numbers.

But blatant phonyfacting aside, political polls are always a good reminder of Mark Twain’s famous saying....

https://en.wikipedia.org/wiki/Lies,_damned_lies,_and_statistics

[proudpapa]

Are you claiming that there is inaccurate information on the internet?

[BigSkyFreeper]

RCP doesn’t poll, the “RCP Averages” poll is just an average of available polls that are taken out there either from independent firms or the media elite.

[Truthsearcher]

Given that the individual polls may all have a distinctively different methodologies, it makes more sense treating each as a individual data point rather than do a weighted average as you suggest.

That 1000 sample poll maybe have been done by a very poor sampling technique, and the 300 sample poll may have a superb sampling technique and is actually much more accurate. If you just do a weighted average you let the much poorer sampled poll dominate the result over the much better sampled poll.

[Riflema]

Agree. The UI leaves a little to be desired.

[eyeamok]

I think it is run by the Quisling Hugh Hewitt.

[outofsalt]

I encourage everyone to lie to pollsters.

[BlueNgold]

While on a strictly mathematical scale that holds water... however

For RCP to perform simple weighted averaging would also fly in the face of statiscal anlysis theory.

To go deeper on only one level is improper. They would need to do whole scale secondary analysis to compensate for adjusted MOE, modeling, and assignment of confidence category.

Its not intended to be a holistic reanalysis. It’s a snapshot. No need to make a mountain out of a molehill IMO.

[Quicksilver]

There would have to be wide variance in both sample size and percentage to make a significant difference.

The weighted average for Trump’s RCP Average (29.5%) shown in the article is:

(.31*507)+(.31*725)+(.25*531)+(.31*395) = 637.12

507+725+531+395 = 2158

637.12 / 2158 = 0.29523633 = 29.5% rounded

RCP could be using weighted averaging. One would have to find an example with wide variances to verify.

[Intolerant in NJ]

Never made much sense to me - doesn’t take into account what the current trend is.....