According to Zogby, Obama had a big night last night (mentioned inside the Reuters write up). I have no idea how that can be when Obama dropped a point.
In the write-up at Zogby.com, there is no mention of Obama having a big night and it instead puts a whole different spin on things. I dunno, is Zogby schizoid?
Who knows. Zogby claims that he is polling 400 people a day. The margin of error on a sample size this small is a whopping 4.9 percent -- meaning literally that whatever numbers for Obama or McCain he comes up with, they could be off by as much as 5 points in either direction (and ditto for undecideds as well).
Zogby's top line numbers are not crazy (although they look suspiciously like Gallup's top line numbers for its "traditional" LV model). But I simply don't trust his methodology.