Bleeding Heartland
An independent website about Iowa politics
 Bleeding Heartland

How Accurate Is Research 2000?

I saw a new Research2000/DailyKos poll for the Georgia Senate race and I was wondering how accurate that firm is.

They have the Martin and Chambliss race to be within 6 points.  That's pretty close considering what margin of error could be.

But then, I remember back to the Research 2000 poll between Greenwald and Latham and think, “Wow, were they way wrong or what?”

Here was the link to the way off Greenwald and Latham race: LINK.

What do you think?  Do they have a good track record other than the Latham and Greenwald race?  They had that race pegged at 5 points when Latham actually won by 22 points.  That's not close, folks.

 

 

About the Author(s)

Bill Spencer

9 Comments

Leave your comment below

  • most of the GA Senate polls

    have the race fairly close. Actually, I’ve seen several showing Martin less than 6 points back.

    The difficulty is in figuring out who will show up to a runoff election on December 2. I’ve heard predictions that turnout will be only 15 percent, in which case no one has any idea who will win.

    desmoinesdem  Sun 30 Nov 9:18 PM

  • like most pollsters, Research 2000 has mixed results

    Mark Blumenthal discussed the accuracy of more than a dozen pollsters in this post:

    http://www.pollster.com/blogs/…

    Nate Silver discussed the “house effect” of various pollsters including Research 2000 here:

    http://www.fivethirtyeight.com…

    I was never given access to any internal polling done by Latham or Greenwald. I assume Greenwald’s campaign never had an internal poll showing the race as close as that Research 2000 poll, because otherwise they would have released it.

    Similarly, I assume that Latham’s internal polls were not consistently showing him 20 plus points ahead, because he would have released those. Also, if he had a comfortable lead during the whole campaign I do not believe he would have started running negative tv ads.  

    desmoinesdem  Sun 30 Nov 9:25 PM

  • Internals

    Here is an article that the Des Moines Register did before the election showing Latham’s internal numbers.

    I’m surprised you missed that, DesMoinesDem.

    http://www.desmoinesregister.c…

    bill-spencer  Sun 30 Nov 10:09 PM

    • those were the internals Latham released

      after his big advertising blitz against Greenwald.

      I am talking about the internals his campaign collected during the summer and early fall. I am quite sure they were not all as favorable as the numbers his campaign released in response to the Research 2000 poll.

      desmoinesdem  Mon 1 Dec 8:03 AM

  • How Accurate is Bill Spencer?

    Bill Spencer’s Predictions

    1. What percentage of the national popular vote with Barack Obama and John McCain receive?

    51 Obama to 49 McCain

    2. How many electoral votes will Obama and McCain win? (538 total)

    286 McCain, 251 Obama

    3. What percentage of the vote will Obama and McCain win in Iowa?

    53 Obama, 46 McCain

    4. What percentage of the vote will Bruce Braley and Dave Hartsuch receive in the 1st district?

    60 Braley, 40 Hartsuch

    5. What percentage of the vote will Dave Loebsack and Mariannette Miller-Meeks receive in the 2nd district?

    50 MMM, 48 Loebsack, 2 Other

    6. What percentage of the vote will Leonard Boswell and Kim Schmett receive in the 3rd district?

    55 Boswell, 45 Schmett

    7. What percentage of the vote will Tom Latham and Becky Greenwald receive in the 4th district?

    58 Latham, 42 Greenwald

    8. What percentage of the vote will Steve King and Rob Hubler receive in the 5th district?

    58 King, 42 Hubler

    9. How many seats will the Democrats and Republicans have in the Iowa House after the election (currently 53-47 Dem)?

    51 Democrats, 49 Republicans

    10. How many seats will the Democrats and Republicans have in the Iowa Senate after the election (currently 30-20 Dem)?

    17 Republicans, 33 Democrats

    11. Which Congressional race in Iowa will be the closest (in terms of percentage of vote difference between winner and loser)?

    2nd CD – MMM & Loebsack

    12. Which Iowa House or Senate race will be the closest (in terms of percentage of vote difference between winner and loser)?

    Hoy v. Sweeney

    13. Nationally, which U.S. Senate race will be decided by the narrowest margin (in terms of percentage of the vote difference, not raw votes)?

    Hagan v. Dole

    14. In the presidential race, which state will be decided by the narrowest margin (again, in terms of percentage of the vote)?

    Ohio

    pistachio  Fri 19 Dec 12:08 AM

Comments