I'm still trying to figure out how folks like Karl Denninger are interpreting a 7-point swing towards Romney as proof that Obama has the election all but sewn up. Especially when Rasmussen shows a similar 46%-46% result. I understand the Electoral College and why some prefer the state-by-state analyses. But the fact remains that Rasmussen nailed the 2008 popular vote while Nate Silver asserted "Obama’s Convention Bounce May Not Be Receding" on the very day that Gallup reported this seven-point move away from him. Either way, it is useful to recall that in polls as in stocks, past performance does not guarantee future results.