Election Models Update

As I write, Sandy is pounding on us, so I will try to get this out before we lose power. During Irene, we lost power for five days. Ugh.

Here's how the pulse index looks:

And the cumulative...

(For an explanation of this, click here.)

It's worth noting there have been three large, definable trends during this campaign. First, a huge upswing for Republicans during the summer, something that went unnoticed by the MSM. This trend was stopped in its tracks by The Democrat convention, which erased fully half the summer's gains for Republicans (this was very much noticed by the MSM.) More recently, there was a Republican uptrend since the first debate. Interestingly, this trend appeared halting at first, but has actually picked up momentum very recently. If Dems can't reverse this soon, they will be very unhappy on election day. Arguably, the most important trend is the last one.

On the other hand, after Romney pulling even a few days ago, Obama has bounced back some in our Electoral College Model, and he now leads 283 to 255:

This is very odd, indeed. Two models, one based on polls, the other based on people making bets, seem to be telling us two different things. More thoughts on this to come.