|
...is that you're using EV calculations to judge the current state of the race, when they tend to lag behind national poll results by a week or even two.
The reason for this is simple: there's a new tracking poll (or six) every day, but not a new poll for every single state. In most cases, at this late stage of the campaign, a survey organization will poll every key state once a week at most (earlier in the race, it might be every two weeks or every month), so any shifts in popular sentiment will take up to a week to show up.
For example, take a look at the Swift-Boat ads, issued at a time when states were being polled, at most, every two weeks. For about that amount of time after the release of the ads, Kerry's EV standing actually went up -- but it was a mirage, because it was still based largely on pre-Swift-Boat state polling data. Once those polls "caught up" with the ads in most states, Kerry dropped like a rock.
Similarly, look at the RNC this year. If you recall, Obama's lead (from his convention bounce) began dropping by about the third day of the RNC, and McCain was ahead in the tracking polls by the weekend, a couple of days after his acceptance speech. But he didn't actually catch up to and surpass Obama in EV until about a week after that, when state polls caught up. (You'll notice a similar phenomenon after the 2004 RNC, only stretched out a bit because there weren't so many updated-every-week state polls back then.)
In fact, it's interesting to take each dramatic event in this year's race, and see what happened to the EV precisely one week after the event in question. Obama started building his lead a week or so after the Wall Street meltdown. He had another jump in the EV about a week or so after his first debate win (notice that it happened just before the "town hall debate").
My suspicion is that, if you were to graph the tracking-poll average versus this EV chart, you'd be seeing a tracking-poll jump around a week before each similar jump in EV, and a similar mirroring of drops.
What this tells me is that the "end-points" of the red and blue lines on the current chart actually reflect Obama's win in the town hall debate (and his almost immediate jump thereafter in the national polls), not anything that's happened since then. We know that the national polls have been tightening. I think we'll have to wait a week to see if that has had any effect in the EV count.
|