Before everyone panics...today's data reflects Kerry's strong day on Saturday that even Rasmussen says was bad data. I show the raw numbers for the last three days to be 44-48-50 Bush, 49-45-48 Kerry. All done over a holiday/weekend period when Rasmussen polling always shows Bush down (usually by as much as 6 points in raw data). The real story won't be known until Thursday. Tomorrow, a great day for Kerry drops off and a bad day for Bush drops off. Expect Bush to move ahead by at least one or two points.
(1) The margin of error suggests that "actual" number lies somewhere within that range, and
(2) These are quoted at a 95% confidence level, which means that you can count on 1 in 20 samplings to be outside the predicted margin.
To make a long story short, Rasmussen got a skewed data set that will roll out of his 3-day running average tomorrow.