Every four years we get the opportunity to become our own statisticians and predictors — peering into the crystal ball and looking for answers as to who will win the next presidential election. As this has been the longest campaign cycle in our nation’s history (just over two years), we’ve had plenty of time to get our thoughts in order.
However, our instincts and beliefs get clouded with the ubiquitous nature of daily tracking polls.
FiveThirtyEight.com has done an excellent job of taking all those polls, regressing them, factoring out bias (to the best it can be done) and presenting a clear picture of what they all say. Yet, regardless of how much work you do to normalize a poll, you’ll hear naysayers shout “that poll has bias built in.” Just this morning, McCain “strategery” man Rick Davis was saying that the polls have too many Democrats (e.g. more than 45%, the historical average) in them, so they’re biased. Is it possible that the country has more people leaning Democratic than Republican for this election, and the poll is capturing true sentiment? Shocking revelation.
Of the prognosticators I know, I’d have to say my buddy David is forming his prediction the best way — rationally. If you haven’t followed his discussion, I highly suggest you do here, here, and especially here.
As for my thoughts, the election goes the way of Virginia, and Obama clears 320 electoral votes.