A. +/- Umpteen Eleventy

Posted on | July 2, 2010 | 22 Comments

Q. What was the statistical margin of error on DailyKos polls for the past two years?

Markos Moulitsas of Daily Kos has sued his polling company, Research 2000 (R2K, for short), alleging that the firm supplied “fraudulent” data. Daily Kos has produced a report that seems to indicate R2K was just making up numbers. I say “seems to indicate,” because:

I’m a journalist, not a statistician; and
This is a pending lawsuit, so we don’t want to convict someone prematurely.

With that in mind, however . . . .

BWAAA-HAHAHAHAHAHA! ROTFLMAO!

DKos (allegedly) got hosed by this polling company, which was (apparently) selling him live phone polls — i.e., not automated “robo-calls” — at suspiciously low prices. And it seems that neither Markos nor his genius readers suspected anything was amiss until statistics wizard Nate Silver analyzed recent polls and discovered that R2K ranked dead last in accuracy among major polling firms.

Republican consultant Patrick Ruffini roughed in some numbers and found something else fishy:

R2K did polling for state-level liberal blogs like Blue Mass Group in the run up to the Massachusetts special election. On January 14, R2K produced a poll showing Coakley with an 8-point lead (while other polls were showing Brown pulling ahead), and in touting the “good” news, Blue Mass Group proudly noted that “Research 2000 does live interviews, unlike robo-pollsters Rasmussen and PPP.” . . .
Did a Massachusetts progressive blog pay more than $6,000 for a top-of-the-line survey when maybe a half dozen other pollsters were polling the race by that point? Really? This begs the question of what Blue Mass Group really paid. And what did Kos really pay?

Exactly. If it seems to good to be true, it usually is and yet it seems R2K was able to find willing clients for its dubious services among liberal bloggers.

Yeah.

Which calls to mind the fact that major “progressive” organizations were among the victims hardest hit when Bernie Madoff’s pyramid scheme finally collapsed.

The moral of the story is, if you’re running a rip-off operation, seek clients who believe in Keynesian economics, anthropogenic global warming, peace through disarmament, Hope and Change and unicorns. We leave you with this cogent economic analysis from a leading progressive intellectual, Peggy Joseph:

Democrats: Because Stupid People Have Rights, Too!

Category: Daily Kos

Comments

22 Responses to “A. +/- Umpteen Eleventy”

proof
July 3rd, 2010 @ 12:21 am

All that hype about what a disaster Y2K was going to be? They just got the first letter wrong! It was R2K!
proof
July 2nd, 2010 @ 8:21 pm

All that hype about what a disaster Y2K was going to be? They just got the first letter wrong! It was R2K!
Rose
July 3rd, 2010 @ 2:07 am

One has to suspect they knew what they were paying for – they were paying for a poll that was to seed public perception – create the appearance that momentum was on their side. People will go with a winner. So they made themselves (the progressives) look like winners.

Now they cry foul, because reality caught up with them.

It’s no different than the Fenton Communications strategy in gerneral, you want to lead people down the path? You put out some phony “scientific” reports, a cute young “expert” telling you of their “findings” and the news media swarms with it like a school of fish. Brain dead fish, but you get the idea.

I am not buying that they were defrauded. If they were, it is funny, but I think they knew exactly what they were paying for. This is cover.
Rose
July 2nd, 2010 @ 10:07 pm

One has to suspect they knew what they were paying for – they were paying for a poll that was to seed public perception – create the appearance that momentum was on their side. People will go with a winner. So they made themselves (the progressives) look like winners.

Now they cry foul, because reality caught up with them.

It’s no different than the Fenton Communications strategy in gerneral, you want to lead people down the path? You put out some phony “scientific” reports, a cute young “expert” telling you of their “findings” and the news media swarms with it like a school of fish. Brain dead fish, but you get the idea.

I am not buying that they were defrauded. If they were, it is funny, but I think they knew exactly what they were paying for. This is cover.
Adobe Walls
July 3rd, 2010 @ 2:52 am

Flush please.
Adobe Walls
July 2nd, 2010 @ 10:52 pm

Flush please.
Estragon
July 3rd, 2010 @ 4:00 am

There isn’t much dispositive in the Politico story, but why should this one be any different? Kos’ “statistical analysis” curiously omitted any actual statistician, nor did they make any particular claim of statistical expertise, experience, or training. The “evidence” they found could just be evidence they don’t know how to read an extract, and they don’t have the raw data.

R2K has been in business a while and makes a living at it. I doubt seriously they are falsifying data unless the client requested it, but they could be using a leftist slant on the weighting. This would account for them being least accurate – someone is least accurate every year because someone has the weakest model. Maybe this year it’s them because they are trying to please Kos, even if he wouldn’t put it in writing.

R2K said there was a big payment due which Kos is trying to get out of. One of the proffers in Kos’ suit claims R2K asked for advance payment for the whole year – maybe they should have held out for that, since Kos is indeed not paying.

Who to believe? A guy who’s been in business through multiple elections, maybe never being the top poll but usually in the pack of the rest, or the leftist deadbeat?

The oldest way to cheat a service provider to use the service but then claim it wasn’t any good and refuse to pay. My bet, that’s exactly what is happening here.
Estragon
July 3rd, 2010 @ 12:00 am

There isn’t much dispositive in the Politico story, but why should this one be any different? Kos’ “statistical analysis” curiously omitted any actual statistician, nor did they make any particular claim of statistical expertise, experience, or training. The “evidence” they found could just be evidence they don’t know how to read an extract, and they don’t have the raw data.

R2K has been in business a while and makes a living at it. I doubt seriously they are falsifying data unless the client requested it, but they could be using a leftist slant on the weighting. This would account for them being least accurate – someone is least accurate every year because someone has the weakest model. Maybe this year it’s them because they are trying to please Kos, even if he wouldn’t put it in writing.

R2K said there was a big payment due which Kos is trying to get out of. One of the proffers in Kos’ suit claims R2K asked for advance payment for the whole year – maybe they should have held out for that, since Kos is indeed not paying.

Who to believe? A guy who’s been in business through multiple elections, maybe never being the top poll but usually in the pack of the rest, or the leftist deadbeat?

The oldest way to cheat a service provider to use the service but then claim it wasn’t any good and refuse to pay. My bet, that’s exactly what is happening here.
Adobe Walls
July 3rd, 2010 @ 4:15 am

Read (can’t remember where) there is some insinuation that Kos played with R2K’s numbers. Isn’t it entirely appropriate that Kos is supplied with misleading numbers? What happened to that plunger?
Adobe Walls
July 3rd, 2010 @ 12:15 am

Read (can’t remember where) there is some insinuation that Kos played with R2K’s numbers. Isn’t it entirely appropriate that Kos is supplied with misleading numbers? What happened to that plunger?
Mark J. Goluskin
July 3rd, 2010 @ 4:18 am

And what, we should be shocked that the Kossacks got had?! Hell no, we should be celebrating! Proving that Rasmussen is still the best in the business. Because they are usually RIGHT! Unlike Kos2000 or what ever the “polling” outfit is called. Oh, one more thing. Ol’ Kos is writing his “American Taliban” based on-polling from his hack pollster! I believe that it will be in the fiction section of Ye Olde’ Bookseller.
Mark J. Goluskin
July 3rd, 2010 @ 12:18 am

And what, we should be shocked that the Kossacks got had?! Hell no, we should be celebrating! Proving that Rasmussen is still the best in the business. Because they are usually RIGHT! Unlike Kos2000 or what ever the “polling” outfit is called. Oh, one more thing. Ol’ Kos is writing his “American Taliban” based on-polling from his hack pollster! I believe that it will be in the fiction section of Ye Olde’ Bookseller.
BlueCollarGal
July 3rd, 2010 @ 5:40 am

Speaking as a person who has several years experience conducting “non-political” surveys via telephone, I feel qualified to say good data collectors are hard to find and we are paid quite well for our abilities.

Speaking as a person who has received numerous political polling calls, I can tell you the questions are ill-conceived, highly suggestive, and presented in a rush by data collectors whose first language is rarely English.

I’m not big fan of robo-polling because it is so easy for a child to press 1 or 2 or whatever is requested, just to mess with the bot. But I’m equally unhappy with the human contact polls that rely on results produced by workers paid by the number of surveys they complete.
BlueCollarGal
July 3rd, 2010 @ 1:40 am

Speaking as a person who has several years experience conducting “non-political” surveys via telephone, I feel qualified to say good data collectors are hard to find and we are paid quite well for our abilities.

Speaking as a person who has received numerous political polling calls, I can tell you the questions are ill-conceived, highly suggestive, and presented in a rush by data collectors whose first language is rarely English.

I’m not big fan of robo-polling because it is so easy for a child to press 1 or 2 or whatever is requested, just to mess with the bot. But I’m equally unhappy with the human contact polls that rely on results produced by workers paid by the number of surveys they complete.
J David
July 3rd, 2010 @ 9:15 am

The second-to-last sentence is really classic “…soul of wit” brevity; very funny…because it’s true.

The wicked prey on venal fools as they are easy pickings.
J David
July 3rd, 2010 @ 5:15 am

The second-to-last sentence is really classic “…soul of wit” brevity; very funny…because it’s true.

The wicked prey on venal fools as they are easy pickings.
steve in tulsa
July 3rd, 2010 @ 12:15 pm

Kos did not get ‘burned’ by his polling company. He told them what to produce and they gave him what he wanted. But then the polling company got caught. Then Kos claimed he was outraged and ‘burned’. That is bull. He was behind the false figuring from the beginning but to save his ass he throws the polling company under the bus and claims he had nothing to do with it.
steve in tulsa
July 3rd, 2010 @ 8:15 am

Kos did not get ‘burned’ by his polling company. He told them what to produce and they gave him what he wanted. But then the polling company got caught. Then Kos claimed he was outraged and ‘burned’. That is bull. He was behind the false figuring from the beginning but to save his ass he throws the polling company under the bus and claims he had nothing to do with it.
Charles Johnson
July 3rd, 2010 @ 1:35 pm

I’ve looked into this and the polling data was accurate. It’s that white supremacist blogger that is making all this up.
Charles Johnson
July 3rd, 2010 @ 9:35 am

I’ve looked into this and the polling data was accurate. It’s that white supremacist blogger that is making all this up.
Robert Stacy McCain
July 3rd, 2010 @ 3:03 pm

GG, you’re missing the point: Nate Silver did that ranking study hoping to *vindicate* R2K’s accuracy. And it was only when their accuracy was demonstrated to be much WORSE than Strategic Vision’s that Markos expressed suspicion that anything was wrong with R2K’s polling — AFTER he had already paid them big money.

According to their own statements, your guys got rooked, swindled, hoodwinked, hustled, scammed and bamboozled.

SUCKERS!
Robert Stacy McCain
July 3rd, 2010 @ 11:03 am

GG, you’re missing the point: Nate Silver did that ranking study hoping to *vindicate* R2K’s accuracy. And it was only when their accuracy was demonstrated to be much WORSE than Strategic Vision’s that Markos expressed suspicion that anything was wrong with R2K’s polling — AFTER he had already paid them big money.

According to their own statements, your guys got rooked, swindled, hoodwinked, hustled, scammed and bamboozled.

SUCKERS!