Kos is changing pollsters, mainly because of Nate Silver’s pollster rankings (more here).
I don’t have much to say about the particulars of this decision, but Kos’ decision to make a significant investment in quality polling, and to be very transparent about it, is a major, positive and often little-appreciated contribution to our national debate. He does more and better polling than a lot of the big media outlets who piss on “blogger parasites”.
Currants
I’m no numbers geek, but I LOVE reading Nate Silver’s site, not least because he makes corrections when he’s wrong, AND he tells you straight, even if the news is that the conservative/Repub candidate’s going to win, not the progressive/Dem candidate. I don’t like reading views that only reflect what I already believe–if I want to construct my own reality, I’ll do that on my own time. :-)
JGabriel
I’d say it’s a good move on Kos’s part, even without Nate’s number-crunching. I hadn’t noticed a particularly partisan lean in R2K’s numbers, but I did notice a lot of variability and inconsistency in their results – not restricted to DKos either.
I’d like to see Kos try Survey USA or Ann Selzer, if she was willing to nationalize her operations. PPP is pretty good too, but they’re already identified as a Democratic pollster, and it would be better for Kos to work with someone that doesn’t have a partisan identification, if merely to avoid the appearance of bias or a conflict of interest,
JGabriel
BTW, apparently Nate did find a Democratic-leaning house effect at R2K. Just wanted to add that clarification and counter-evidence, since I noted above that I hadn’t noticed one.
.
PanAmerican
So when his political bias ran into reality he didn’t plug his ears and start yelling louder about hippie punching? Stupid, corporatist, Obamabot sellout. Nobody in the veal pen avoids becoming Veal Scaloppini, man.. nobody.
Markos’ passion is clearly the horse race stuff (I find the rest of the front page a delightful combination of mediocre writing and editorial incoherence) and I agree about the value of the polling. I just wonder if the results weren’t baked into the nature of the gig. A High volume, high profile, partisan customer
mayis going to generate biases regardless of who is doing the polling.I don’t get a sense of whether he tried(and failed) to address any issues with R2000 or if it was an impulse whack on account of the AR results and Nate’s rankings.
Martin
@PanAmerican:
He did, actually. R2K changed how they were doing their polls based on input from Kos (presumably based on feedback from Nate) – and made them more balanced. This was 4-6 months ago, IIRC. I’m not sure they’re balanced enough, though. Too bad, as he was commissioning polls that nobody else was running, so I found myself running back to his quite frequently.
I’m hoping he picks up SUSA – I’ve always been impressed with the accuracy of their polls. Not perfect (nobody is) but my off-the-cuff sense is that SUSA has little if any bias in their errors.
Joel
Nate does good work.
ThresherK
I have no opinion about Nate Silver until his ESPN.com forecasts win me the World Cup pool :-).
Really, though, this whole “poll the pollsters” stuff has shown how much hackitude is created by Zogby and Rasmussen. If that’s its only reason to exist, that’s reason enough.
randiego
The polling that Markos does is one of the strengths of that site. I’m not interested in an echo chamber, I want reality!
BombIranForChrist
I am a pretty big critic of Kos, who I actually think is pretty screechy. I think he’s often right about a lot of things, but his approach is often obnoxious, too. There’s just too much of this: “Well, I’m just going to take my ball and go home”.
UPDATED:
Hmm, scratch that. It appears that his current pollster isn’t really that wrong after all. As Kos admits, there is not a lot of difference between it and “better” polls.
Right now it is June. November is not far away at all. To just dump all your polls and trend data this close to a big mid-term doesn’t seem the right thing to do. Trends are so critical, that I am not a little skeptical. What, he will have 3 months of trend data with the new pollster only so he can shave a few percentage points off accuracy errors? I wouldn’t have done it.
Still, good for him. I never really doubted his integrity, only his methods.
BombIranForChrist
Sigh, I didn’t get through editing my last post before the editing buzzer sounded, so it is probably more nonsensical than usual, even for me.
My point:
1. Kos is screechy.
2. But he has integrity.
3. But he should have waited until after this polling season to start over. Tiny increases in accuracy don’t warrant dumping a pollster this close to election.