Barack Obama had another strong day in the national tracking polls, increasing his advantage in the Rasmussen Tracker to +6, and in the Gallup Tracker
to +5. Indeed, four of the five tracking polls are now in agreement
that Barack Obama's lead is in the 5-6 point range, with Battleground
dissenting and putting the race at McCain +2.
You should bear in
mind, however, that these polls reflect the pre-debate state of the
race, as the overwhelming majority of the interviews for the these
tracking polls took place before last night's debate was completed. A
variety of reactions to the debate seem possible to me, including (in
rough order of probability): i) a small gain for Obama; ii) no
effective change; iii) a larger gain for Obama; (iv) a small gain for
McCain. The reaction in the horse race polls in the days following the
debate do not always match the overnight flash polls, as opinions about
the debate may change once filtered through the lens of the media.
However, since Obama won or tied essentially all objective evaluations
of public opinion about the debate, material gains for McCain appear
unlikely.
At the state level, we have a relatively uneventful day:
The only poll that really stands out is ARG's McCain +3
in Colorado. Since (i) a lot of good pollsters were in the field last
week in Colorado and showed Obama with a decent-sized lead, and (ii)
ARG is not a good pollster, this does not really affect our model's
opinion that Colorado leans fairly strongly toward Obama. Otherwise,
Iowa looks good for Obama ... Missouri is going to be tough for him,
even if he has a good night overall ... these are things that we knew
already.
Obama, however, did tick upward a couple of points in
our win percentage estimate, and is now given a 78.5 percent chance of
winning the election. Our more eagle-eyed observers may notice that
this improvement occurred even as our popular vote projection barely
changed. There are two reasons to explain the discrepancy: (1) Obama
got a comparatively "bad" polling result in California (+10 from
SurveyUSA), and California has a fairly substantial effect on the
popular vote but almost none on the electoral college, and (2) We are
now close enough to the election where every day that goes by without
McCain making gains in the polls makes him marginally less likely to
win. So Obama's win percentage will be ticking up by perhaps half a
point a day based on inertia alone.
--Nate Silver