Free Republic Browse · Search News/Activism Topics · Post Article

To: MrChips

Rasmussen has been increasing Democratic weighting because Republicans aren’t answering the poll calls. According to the rules of polling science, he has to assume that his sample is still representative, and more people really are identifying themselves as Democrats.

(We have good reason to believe that Republicans simply aren’t answering the phone, thus skewing the people who do answer towards the Democratic, but it’s impossible to prove that.)

Let’s say that the breakdown is R33/D33/I33.

If 100% of Rs vote for Romney and 100% of Democrats vote for Obama, and if Independents went 50-50, then the result is 50%R-50%O.

Now let’s assume independents had a 16-point swing and all else were equal.

You’d have Romney = 0%D-100%R-58%I and Obama = 100%D-0%R-42%I. Calculate the total and it’s Romney=52.7% and Obama=47.3%. That’s “only” a 5.4% point swing.

Of course, the reality is that independents are not 33% of the voting electorate and Romney was already leading with them; I am simply positing a “best-case” scenario for independents to speak.

41 posted on 10/06/2012 7:20:51 AM PDT by Cruising For Freedom (Don't be the proof that MSM PsyOps works.)

Thank you for all the math! I am a Latin teacher, not a math guy! But, from your numbers, I can now extrapolate a general rule for myself: that in a perfect 33/33/33 model with parity between D & R, a 16 point advantage among Independents would only mean a 5 or 6 point move. So a sample of 20% Independents would move things only 2 or 3 points. So, with a +4 Dem sample (which well may not hold), a +2 lead for Mitt is pretty damned good!

57 posted on 10/06/2012 7:34:30 AM PDT by MrChips (MrChips)