Gallup is Very Upset at Nate Silver

[octoberlib]

http://www.salon.com/2012/11/13/gallup_is_very_upset_at_nate_silver/

Did Gallup just blame Nate Silver for ruining the art and science of polling?

You don’t have to read too far between the lines of a statement from Gallup’s editor in chief, Frank Newport, published on Friday, to get that impression.

Newport first attempts the formidable task of defending Gallup’s polling accuracy during the 2012 campaign. Perhaps he was anticipating Silver’s Saturday column, which labeled Gallup the most inaccurate pollster of all the firms that measured voter sentiment this year. But Silver was hardly alone in wondering why Gallup regularly reported numbers much more favorable to Romney than anyone else in 2012. We deserve an explanation a little less lame than Newport’s: what’s the big fuss? Gallup wasn’t really off by that much.

But then it gets interesting:

"But some of this will result from a variant of the venerable “law of the commons.” Individual farmers can each made a perfectly rational decision to graze their cows on the town commons. But all of these rational decisions together mean that the commons became overgrazed and, in the end, there is no grass left for any cow to graze. Many individual rational decisions can end up in a collective mess.

We have a reverse law of the commons with polls. It’s not easy nor cheap to conduct traditional random sample polls. It’s much easier, cheaper, and mostly less risky to focus on aggregating and analyzing others’ polls. Organizations that traditionally go to the expense and effort to conduct individual polls could, in theory, decide to put their efforts into aggregation and statistical analyses of other people’s polls in the next election cycle and cut out their own polling. If many organizations make this seemingly rational decision, we could quickly be in a situation in which there are fewer and fewer polls left to aggregate and put into statistical models. Many individual rational decisions could result in a loss for the collective interest of those interested in public opinion.

This will develop into a significant issue for the industry going forward".

Priceless.

[durablend]

1. Translation...

"Romney could've WON if it weren't for you MEDDLING KIDS!!!!!"

[PsychProfessor]

2. I saw Newport once in person and met him.

At the time he was aglow with how wonderful GWB was. Seriously. I was not impressed and remain unimpressed. If you can't stand the heat....

[UCmeNdc]

3. Gallup needs to get their LV model more in line with the actual LV that vote

Is that too much to ask for?

[octoberlib]

7. Since the RV model is always more accurate , I think they should dump the LV entirely nt

[UCmeNdc]

15. You are right

Gallup RV was not that far off. It was their LV that seemed to poll the outer space zone.

[alcibiades_mystery]

4. I vote for Gallup to stop conducting polls!

[Jeff In Milwaukee]

21. That's 100% in favor...until we unskew that poll, of course!

[courseofhistory]

5. Boo hoos! Get out of the business, Gallup! n/t

[democrattotheend]

6. In principle, I don't completely disagree

Regardless of Gallup's mistake in using too tight of a likely voter screen, their methodology for conducting the poll is pretty sound, and I don't want to see polls like that become obsolete in favor of cheaper robopolls or online polls. I still think there is value in having a reputable pollster who uses live calling, including cell phones, and uses a large sample. I hope this cycle doesn't render live polls like Gallup's obsolete.

I don't see his point in attacking Nate Silver, but he is right that Nate would be out of a job if polling organizations stopped making the investment in conducting polls. But I don't see it as an either/or - both polling and aggregation/analysis of polls are useful.

[Hutzpa]

9. In that case they need to change their method of sampling

start producing facts instead of made up lies.

Try sampling every region and ethnicity with exact numbers as oppose to rounding up numbers and lastly, fuck Gallup for being biased toward republicans try and be neutral.

[octoberlib]

11. Well , aren't most polls conducted by marketing firms?

Most poll aggregators are political scientists or in Nate's case , statistical analysts. I can't see Gallup quitting polling.

[geek tragedy]

13. The problem is that Gallup isn't taking accountability for an obviously broken LV model

that has produced absurd results for the past few election cycles.

They're sitting there and pretending that 50-49 Romney when in fact it was 51-48 Obama was a pure sampling error problem--when in fact their RV results called the race perfectly.

[unblock]

8. translation: we're gonna start charging people like silver (or the nytimes) a fee.

and we'll sue if we have to. we'll figure out a way to legally enforce it.

[Hutzpa]

10. Ain't gonna happen. nt

[Cha]

12. Of course, gallup would take no responsibility for ruining

Themselves!

[LisaL]

14. Not off by that much means Gallup projected the wrong candidate to win.

[Iggo]

28. Yeah, so...

...of the two candidates with a chance to win, they were only off by one.

Heckuva job there, Gallup.

[avebury]

16. Well isn't that just too bad! If Gallop can't change

with the times they need to just go away.

[fugop]

17. Bigger problem that Gallup SHOULD complain about

The media. Gallup has a right to conduct polls. They even have a right to their take on who likely voters will be. The problem was that the media constantly pimped the Gallup poll's likely voter model without explaining what it was. It misled viewers constantly into believing, "The president is losing! Romney has all the mo!" Had the lazyass media reported both numbers (RV and LV) and explained to voters the disparity, it wouldn't have been so incredibly dishonest.

But I don't excuse whining Gallup because he knew damn well that the press would breathlessly obsess over the LV numbers and ignore all else. So he can suck it now.

[octoberlib]

18. I totally agree.

Time after time I saw the media cherry picking the polls that were favorable to Romney or showed the race tied.

[DemocratSinceBirth]

19. FOX Democrat & Clown ,Bob Beckel ,Said The Election Was Over When O Was Down 52-45 the Gallup Poll

When we all knew it was garbage.

[Third Doctor]

20. Nate made them look like idiots

and they are pissed about it. It's their own fault as usual.

[ItsTheMediaStupid]

27. Their own LV model made them look like idiots

Nate just saw a outlier, called it an outlier and discounted it appropriately in his model.

Nate was like a professor I had once. He laid out the criteria he used and how he calculated the grades. Then he said we make the grade based on our work. He just kept score.

[NHDEMFORLIFE]

22. What a bunch of double talk

If Newport actually believes in the gibberish quoted here, Gallup is going to be going belly-up before the 2014 mid-terms. And I mean its entire operation, not just that which clings to archaic methods of measuring political activity. I can't believe any potential new clients are going to give Gallup the time of day after its awful performance this year.

[Awsi Dooger]

23. These graphs are devastating to Gallup

Check the trend graphs at this link, the long demonstrated shift of the electorate as opposed to where Gallup pretended it would be in 2012:

http://themonkeycage.org/blog/2012/11/07/galluping-away-from-the-herd/

As I posted a few days ago, Frank Newport is a jackass. That's been demonstrated every time his polling comes under scrutiny.

Frankly, Gallup and others were spared for a long time. The marvel of sophisticated aggregation is not that it's happening now, but that it took so long to take hold. I started betting politics in the early '90s and had great success even though I knew my Excel models were not overly sophisticated. They were a hobby between sports betting. I'd devote a few hours per week. The betting odds at that point were based on guesswork and punditry. An oddsmaker would decide -160 favoritism sounded good so he'd throw it up there. I knew I had a huge edge based on superior knowledge of a few key demographics, plus the partisan tendencies of each state. Nate and others like him have ruined the value aspect of political betting but it's a joy to watch others howl in protest as he earns more praise and mystique than they could threaten to manage.

[sarcasmo]

24. One more reason to love Nate Silver.

[WeekendWarrior]

25. Dear Gallop: Fuck. Off.

[Jennicut]

26. Gallup screwed up by overestimating the amount of the white vote. That is why their polls sucked.

And will continue to suck. Dear President of Gallup: Demos. Look them up sometime. Look at the electorate every year. % of White voters going down. Thank you.

Idiots.

[Incitatus]

29. Sorry the election wasn't stolen, Gallup.

Then you wouldn't look bad.

[SimplyMarie]

30. Really?

That's a pretty funny statement coming from the organization that rated the worst in the accuracy of their polling during this election cycle.

Nate Silver is a genius...and hopefully we all learned a lesson when it comes to trusting the validity of polls.

[Quantess]

31. Sour grapes!