Why can’t they give a meaningful poll?
A recent poll claims that in a head to head matchup that Barack Obama has a better shot at beating John McCain than Hillary Clinton. The poll is a nationwide poll of likely voters. Barack beats McCain 48-41 while Hillary and McCain are tied at 46. I would guess that one reason for this is that while conservatives are undecided on if they could actually come out and support McCain against Obama, they are probably sure that they would vote for anybody against Hillary. This is one of the things that makes me lean towards Obama (though I still have yet to make up my mind).
The problem with the poll is this. It does not actually give meaningful results about what the actual election results would be! We do not choose our President based on who gets the most votes nationwide. The President is chosen via the electoral college. So telling me who gets the popular vote doesn’t give me a meaningful answer with such a close race. Why can’t the polling “experts” take a poll in each state (and in each congressional district in states that award electoral votes for winning congressional districts) and then determine who would get the electoral votes based on those poll numbers.
February 8th, 2008 at 7:03 pm
Doing fifty polls is much more expensive than doing one poll. Plus, there are undecideds and states falling within the margin of error probably shouldn’t be designated one way or the other– it’s a statistical tie. Also, third party nominees aren’t yet known, adding another small variable.
February 8th, 2008 at 7:30 pm
You know, I was going to acknowledge that doing 50 polls is more expensive. Of course, i can buy a $50 fridge and a $2500 fridge, which one is more likely to give me better results? The answer? The one you paid 50 times more for.
The other thing to consider is that certain polls do not really concern me. I think I can guess which party will take MS and which party will take CA. Tell me who does better in PA, OH, and FL (the swing states). Perhaps add some of the other close ones from 2004 as well. That way, you could make some logical assumptions about some states while giving meaningful results about other ones.
And for the third party candidates, they would probably fit under “other” (unless a certain, well funded, ind. candidate enters the race).
Anyway, if I could know that if Obama vs McCain would lead to Obama taking Louisiana (because of Republican apathy for McCain and extreme turnout from minorities for Obama) then I would absolutely cast my vote for Obama tomorrow. Polls telling me that Obama beats McCain straight up in the popular vote that don’t say he gets enough electoral votes to win do not help me make my decision.