RYAN PLUS ONE – AT 9:51 A.M. ET:  There are no reliable polls thus far that gauge the immediate impact of the Ryan choice.

Rasmussen has just reported his tracking results for Sunday morning, but they are based on polls completed before Ryan was named.  Rasmussen shows no change from yesterday, with Romney two points up on Obama.

We won't have early public responses to Ryan until early in the week.  My own sense is that the Republicans handled the Ryan rollout reasonably well, but that the response of Democrats was fuzzy and somewhat less intense than I'd thought it would be.  Buy remember, it's a weekend.  The heavy guns won't start firing until tomorrow.

A word about these polls:  We've been subjected to major yapping in the last week about three polls showing Obama expanding his "lead" over Romney.  And yet, wait a minute, the two major tracking polls, Gallup and Rasmussen, show an exceptionally close race or even a dead heat, in Gallup's case.  How so?

Polls can only be as accurate as the choices made of who to poll.  If you polled only in upper Manhattan in New York, Obama would get more than 90% of the vote.  If you polled in Oklahoma, the result would be dramatically different.

Some of the polls that show a wide Obama lead are oversampling Democrats because they are using old models of the electorate.  That doesn't mean there hasn't been a pro-Obama trend in the last few weeks.  I don't know.  No one actually knows for sure.  I've found that the tracking polls tend to be closer to the electorate and to correct their errors more quickly.  The campaigns themselves have internal polls, and I don't see much gloating on the Obama side.  The gut feeling is that the race is very close.

Which polls will turn out to be the most accurate, when the post-election study is done, is impossible to predict.  Former standing does not guarantee current success.  The most accurate polls in the 2008 presidential election turned out to be Rasmussen and the Pew poll.  Yet, they currently show widely differing results.   Gallup did poorly in 2008, but Gallup and Rasmussen are today remarkably close.

However, even polls that oversample one group or the other can have value in showing trends among sub-groups.  It is clear that Romney is having a problem among women and independents.  If he doesn't address the problem, he will lose.  It is clear that Obama, with all his bluster and leadership position in some polls, cannot get his approval rating up to 50%.  If he can't do it be election day, Romney could squeak in, based on what the history of elections shows us.

Hang on.  It's very early.

August 12,  2012