Gallup's Own Polling Numbers Indicate an Obama Landslide, Not a Romney Lead

I remember every political pundit I listened to just 2 or 3 months ago, telling me that all we had to do was watch the President's approval numbers. If he was 50% or above, he couldn't be beaten.

Occasionally, as in 2004, the incumbent could be just slightly under 50% and still win (Bush at 49.7%).

Gallup's explanation in Jan. 2012
http://www.gallup.com/poll/152051/Obama-Faces-Challenging-Election-Climate.aspx

Today (10/24), Gallup has Romney with a 3 point lead nationwide with Likely Voters. Even accounting for the fact that Obama's approval numbers are with Registered Voters, his 53% to 42% approval-disapproval spread this late in the game is more suggestive of an Obama electoral vote landslide than a 3 point Romney lead.

I follow these numbers very closely. I remember what all the pundits say. These numbers, from the same pollster on the same day interviewing the same people, make no sense at all, even using the pollster's own reasoning.

Link;
http://www.gallup.com/home.aspx