Two thoughts on polls…

1) The guys at 538.com examine the “shy Tory” theory, and wonder if the partisan composition of the electorate really changed as much between 2004 and 2006 and today as we thought, or whether some Republicans just didn’t want to tell pollsters that they identify with the unpopular GOP… If the face of the Republican Party is Tom DeLay, Dennis Hastert, Duke Cunningham, Mark Foley, etc., right-leaning voters no longer identify themselves as members of the party; once the face of the party is McCain and Palin, they say once again, “yes, that’s me, I stand with them.”

2) Obama’s peak in the tracking polls seemed to come a bit later than teh Democratic convention, roughly when the coverage of Hurricane Gustav was at its peak. Katrina flashbacks, perhaps? People see New Orleans preparing for a hurricane, they think of Katrina, they’re reminded of how angry they were at President Bush over it, and they drift towards the Democratic candidate? And does Ike, or further developing tropical storms, mean that Obama’s best shot for a big win is a late October hurricane hitting the Gulf Coast?