Rasmussen Polls vs. 2006: Final Senate Results

Rasmussen Reports ^ | 11/10/06 | Rasmussen Reports

[NYC Republican]

Take a look at Ras' homepage... Look at the closest races, or if you prefer, all of the races they polled. Pretty darned accurate. http://rasmussenreports.com/

[edsheppa]

That's ridiculous. ... So much can change in two weeks, and so much did.

You are completely missing the point. Election polls are valuable only to the degree they are predictive.

[NYC Republican]

??? please 'splain

[edsheppa]

I will try to explain. For many years now it has been the practice during election season for the media, and politically interested people such as we here at FR, to focus on election matchup polls for many weeks and months ahead of the election. This is all wasted effort as the predictive power of these polls is not very good. OTOH, polls taken just before an election are often quite good. But then what's the point there? To the extent that they're accurate, the election being predicted has already been decided.

Now election matchup polls, early and late are probably somewhat valuable to the political pros. National parties see problems and opportunities and can move resources around to their advantage. But their just not so interesting in their own right. And other kinds of polls are probably much more useful.

I'd say in general that election matchup polls have limited tactical utility. It's much more important to have a good strategy. The Ds this past season have had the better one and were aided by many factors. The R strategy was not very good. I hope they can find their way. It's ridiculous for the center right party to lose in a center right nation.

And I apologize for not being clear earlier.

[NYC Republican]

OK, makes sense. Thanks for clarifying