Today's Polls, 10/4

With the first set of tracking polls out to incorporate at least one full day of post-debate interviewing, there is no indication that John McCain and Sarah Palin have made progress in closing their gap with Barack Obama. In fact, Obama ticked upward in three of the four national tracking polls that published today, although he lost a point in Rasmussen. In addition, Ipsos/McClatchy has come out with a poll showing that the debate moved undecideds slightly toward the Obama ticket, confirming the results of most of the snap polling conducted on Thursday evening.

Beyond that, this is the slowest polling day that we've had in some time. The only fresh state poll out is in Maine, and it's actually a pretty good one for John McCain: Rasmussen has him trailing there by just 5 points, little changed from their survey two weeks ago that had him down 4. Although several recent polls have shown Maine tightening, our model will need additional evidence before it concludes that Maine is any sort of toss-up state, as Obama's polling has been strong there all year and as things are going well for him in the Northeast region (as they usually do for a Democrat). Nevertheless, Maine is a very cheap state to and compete in, so our return-on-investment index likes it as a long-shot play; the McCain campaign's decision to commit resources there looks wise.

Finally, just a couple of bits of housekeeping. Firstly, we have corrected the results from the Elon Poll in North Carolina that we published yesterday. We had originally listed the poll as showing a 2-point Obama lead; however, these were the results of a question asking which candidate would handle the economy better, and not the Presidential horse race. The horse race numbers are there, but require a bit of digging to find; they show a virtual dead heat (Obama 38.9, McCain 38.8).

Also, our poll detail chart will now list state tracking polls in a different way. Instead of seeing this:



...that is, each day's results from a tracking poll listed individually (but with very small weightings because few new interviews are conducted each day), you will instead see this:



...that is, a weighted average of the tracking poll's results consolidated onto one line (and highlighted in a salmon color).

This is purely a cosmetic/presentation issue; behind the scenes, our model will continue to treat each day's results from a tracking poll as a separate data point. But it gets very messy to show things to you this way.