Quantcast

Comments about ‘Robert J. Samuelson: Storm brewing over how opinion polls are taken’

Return to article »

Published: Tuesday, Oct. 30 2012 12:00 a.m. MDT

Comments
  • Oldest first
  • Newest first
  • Most recommended
Roland Kayser
Cottonwood Heights, UT

Nate Silver is the pollster who most accurately predicted the last election. He missed only one state, Indiana, a reliably Republican state that went for Obama by the tiniest of margins. His prediction for this election is Obama 296 electoral votes, 26 over the magic number. He also says there is a 7% chance that Romney will win the popular vote while losing the electoral college.

procuradorfiscal
Tooele, UT

Re: "Nate Silver is the pollster who most accurately predicted the last election"

Actually, Nate Silver is not even a pollster. The former Daily Kos commentator only averages and massages other people's polls. He rarely uses or heavily weights Rasmussen or Pew.

Interestingly, Fordham University's rating of pollsters after the 2008 elections found Rasmussen and Pew to be the closest to accurate.

Today's reputable poll results?

Rasmussen -- Romney 49%, Obama 47% [50% Romney to 48 Obama in Ohio].
Pew -- Romney 47%, Obama 47%

Hmmmmm.

LDS Liberal
Farmington, UT

procuradorfiscal
Tooele, UT

Rasmussen...to be the closest to accurate.

==========

Says the Ditto-head.

Rush Limbaugh is the only one who quotes and repeats that about Rasmussen on a daily basis.

I'm sure there's some $ involved there to say that.

procuradorfiscal
Tooele, UT

Re: "I'm sure there's some $ involved there to say that."

Is that an intentional, or unintentional slander of Fordham University academics?

Roland Kayser
Cottonwood Heights, UT

To procuradorfiscal: You are correct, Nate Silver is a statistician, not a pollster. Still, the fact remains that he called the last election better than anyone. We'll see this time.

The Skeptical Chymist
SALT LAKE CITY, UT

To profiscador: Take a look at Nate Silver's site and tell me again that he doesn't use Rasmussen. It is listed there in all the swing state analyses, along with the weights. He has explicitly said that he doesn't exclude any polls. If you wish to discredit his methods, you can't do this with untrue statements. He certainly includes Rasmussen in his analyses. If you think he doesn't assign Rasmussen enough weight, then explain why you think so. If you think his analysis is invalid, explain why the betting markets are generally in agreement.

It is easy to dismiss someone's work when it doesn't agree with your preconceived ideas. That just demonstrates how tightly you hang on to your preconceived ideas, nothing more.

RedShirt
USS Enterprise, UT

To "Roland Kayser" and how many other elections has Nate Silver been correct at predicting? All you have established is that he guessed right 1 time. Like the old saying goes, even a broken clock is right twice a day.

If he was right once, that proves nothing. Can his methods be used to model previous elections. Tell us more.

You might as well use the myth surrounding the Redskins, and if their correlation to who wins the Presidency.

procuradorfiscal
Tooele, UT

Re: "It is easy to dismiss someone's work when it doesn't agree with your preconceived ideas."

I guess I'll just have to take the word of one who has WAY more experience than I, on how easy it may or may not be to dismiss someone who doesn't agree with preconceived notions.

But, anyone looking at Nate Silver's Daily Kos work, as well as what he's done since then, can attest that it's absolutely true that he "rarely uses or heavily weights Rasmussen or Pew."

one vote
Salt Lake City, UT

All polls should be suspect in light of the fact political leaders, waist deep in money, use biased polls to attempt to influence elections.

Rifleman
Salt Lake City, Utah

Re: LDS Liberal Farmington, UT

For a while you were giving us updates on the presidential electoral college vote count. I don't suppose the death of our ambassador to Libya has effected the number. Who knows, perhaps the undecided voters will give him another chance. Then again, perhaps not.

Rifleman
Salt Lake City, Utah

Re: LDS Liberal Farmington, UT

Come Nov. 7th there are going to be some unhappy people wondering why the pollsters got it so wrong. The Obama re-election committee would like us to think the race is neck to neck but it isn't.

Obama is toast and you can see it in his behavior.

JoeBlow
Far East USA, SC

Polls are one aspect. But, if you want to know the real odds, look at those with money at stake.

Look at the betting odds.

As of right now,

bet $1000 on Obama, win $384
bet $1000 on Romney, win $2000

It is all about 1 or 2 swing states. Nothing else matters.

The Skeptical Chymist
SALT LAKE CITY, UT

To procurafiscador:

Time will tell whether or not Nate Silver's model is accurate. Perhaps we should talk again after Tuesday.

procuradorfiscal
Tooele, UT

Re: ". . . we should talk again after Tuesday."

Yeah, let's.

to comment

DeseretNews.com encourages a civil dialogue among its readers. We welcome your thoughtful comments.
About comments