The Zogby tracking poll now shows Bush with a miniscule lead against Kerry, and a dead heat overall. Yesterday, the Zogby poll showed Kerry ahead by three. Obviously a daily tracking poll is showing will show a great deal of volitility, but Zogby’s results seem very odd. A 3-point shift in 24 hours would seem to me to be a sign of something off in methodology – then again, it could also be indicative of a real Bush bounce (the poll is the first Zogby poll that is entirely post-debate.
Of course, this poll also is a Saturday/Sunday/Monday poll, which means that a subsequent weekday sample could very well show another increase for Bush, pushing him ahead of Kerry. However, for Republicans it’s still wise to campaign as though Bush were 10 points behind…
Rasmussen’s daily tracking polls regularly fluctuate four points day-to-day. None of them seem particularly reliable to me.
Tracking polls generally aren’t, although both Rassmussen and Zogby are using three-day rolling samples, which tend to be fairly correct. I think the reason that the Zogby poll is showing such a shift is that they got rid of a weak day for Bush (Friday before the debate) and added a day where Bush did much better (Monday). I don’t have the daily data for Zogby, but I’d guess you’d see a very narrow Kerry lead through the weekend and a big Bush lead on Monday – we’ll see if that pattern holds true in tomorrow’s sampling.