Battleground Final 50-48 Obama

[Sam Spade]

  This is not Battleground's final poll results.  The final results were 49%-44% in favor of Obama.

  This is the Tarrance Group (GOP) portion of Battleground's final projection of the results.  Obama 50.2%, McCain 48.3%, other 1.5%

  For what it is worth, the other half of Battleground, Lake (Dem) is projecting the results as Obama 51.5%, McCain 46.5%, and other 2%.

  We will soon see who is closer.
 
 I have to laugh since Battleground is claiming that they were most accurate in 2004 when all along I thought IBD/TIPP was making that claim.  Guess it all depends on what a person's definition of 'most accurate' is.  LOL

Ill-Ind

IBD/TIPP was the most accurate in 2004, not Battleground, mainly because the final margin was Bush 50.7, Kerry 48.3.  Still, Battleground was quite good in that election, just not as good.  And in truth, that was Ed Goaes's prediction, not Celinda Lake's, who predicted a Kerry victory.

The difference between Ed's numbers and Celinda's numbers in this election have to do with allocation of undecideds.  Celinda thinks they will break evenly, Ed thinks they'll break 4-1 McCain.  We'll see.

[Sam Spade]

  This is not Battleground's final poll results.  The final results were 49%-44% in favor of Obama.

  This is the Tarrance Group (GOP) portion of Battleground's final projection of the results.  Obama 50.2%, McCain 48.3%, other 1.5%

  For what it is worth, the other half of Battleground, Lake (Dem) is projecting the results as Obama 51.5%, McCain 46.5%, and other 2%.

  We will soon see who is closer.
 
 I have to laugh since Battleground is claiming that they were most accurate in 2004 when all along I thought IBD/TIPP was making that claim.  Guess it all depends on what a person's definition of 'most accurate' is.  LOL

Ill-Ind

IBD/TIPP was the most accurate in 2004, not Battleground, mainly because the final margin was Bush 50.7, Kerry 48.3.  Still, Battleground was quite good in that election, just not as good.  And in truth, that was Ed Goaes's prediction, not Celinda Lake's, who predicted a Kerry victory.

The difference between Ed's numbers and Celinda's numbers in this election have to do with allocation of undecideds.  Celinda thinks they will break evenly, Ed thinks they'll break 4-1 McCain.  We'll see.

are the other numbers really true, how accurate their projections were in 1996 and 1992?

They were absolutely dead-on in 1992 and 1996.  No one did better in 1992 and only maybe Zogby did better in 1996 (where he got his rep, unfortunately).

[Sam Spade]

Well...one question then (especially to Sam)

Considering how accurate this poll has been in the past...why the heck have we been paying attention to Rasmussen and Gallup the entire election cycle?

I mean...a win's a win, I guess....so I'm not too worried about that.....but I'm just wondering if you have reason to expect a higher margin than Battleground suggests?

Probably because Battleground stuck to their original weighting the first week and then overhauled it soon afterwards (specifically getting rid of the party ID weights).  Then it started to get kinda bumpy.

In other words, rapid change makes people nervous about the quality.  Understandable.

We believe the other polls because we have been trained to believe that the majority of polls are likely correct, which they probably are.