OxMethod Update

It's been a while since I've updated my polls (weighted by age and accuracy), but there's a doins' a transpirin'!  The latest map is below, along with a chart showing movement.  Most of the movement is obviously toward Obama.  In the national polls, Obama has now moved to a 5.7 point lead, 49-43.3%.  Interestingly, when you weight state polls by the size of the 2004 electorate in each state and then average them (which should add up to the national electorate) you get a narrower Obama lead of 48.4-45.6%.  Not sure what to make of that, but it is out there.  In previous iterations, the two measurements have generally been within a point of each other.

This has utterly destroyed McCain's electoral advantage, and moved him to a 185-353 advantage to Obama.

The other interesting observation is how uneven the movement has been.  Some states have actually shown some movement toward McCain in the last week-and-change -- SC, ME, MT, CA, etc.  Other very important states have not moved at all. IN, AZ, CO, NM, MN are all within a point of where they were on September 22. Some states, though, such as TN, TX, NC, FL, PA, and WA, have seen movement toward Obama of more than five points.

Even given that the model weights (very lightly) polls taken as much as two weeks ago, this is surprising, because all of the polling included in the dataset is now done after the AIG/Lehman Bros. collapse.  This discrepency therefore can't be explained simply by the "lag" that's created by older polls.  I don't have any theories as to why this would be the case, but if anyone has any, I'll be happy to hear them!

One other piece of good news -- Although McCain is down, enough states are close that he isn't out.  He's within at least 3 points in states totalling 271 electoral votes.  It's never good to have to run the table to win, but hopefully once this bailout stuff passes and Bush recedes into the background, some of these states will come back over "on their own."