But since not all polls use likely voters (actually most in the RCP lineup do not), it skews the entire result.
You know, not that I take you seriously anyway when you talk about polling, but this sh
it just makes it too easy. Did you even bother to look
at RCP before making this statement? I sure hope not because the RCP national average right now [ Screen shot since the page will change
] consists of eleven polls, nine of which use a Likely Voter screen. Gallup uses registered voters until closer to election day but has a very large sample size which helps mitigate the variance. The other poll is a IBD/CSM poll.
In fact, if you eliminate the Gallup and IBD/CSM numbers you wind up with a higher margin
for Obama of 3.11 rather than 3.0 and a higher voter share
of 48.77 versus the 48.5 listed.
So, congratulations. You've totally schooled us and shown us that the lack of a LV screen in the RCP average is skewing the results... in completely the wrong direction. Edited, Sep 18th 2012 10:54am by Jophiel