HOW TO OVERHYPE: John Zogby should know better. This is how he reports his latest poll of 2004 Democratic presidential hopefuls:

"North Carolina Senator John Edwards has surged into a tie for second place among rivals for the 2004 Democratic nomination, latest Zogby America Poll results show.

Among likely Democrats nationwide, Connecticut Senator Joseph Lieberman leads with 11%, followed by Massachusetts Senator John Kerry and North Carolina Senator John Edwards, both at 9%. House Minority Leader Richard Gephardt is next at 8% and Senate Democratic Leader Tom Daschle follows at 7%.

In July, Edwards was a distant seventh place among 2004 Democratic hopefuls with 2% support among likely voters.

The poll of 432 likely Democratic voters, was conducted Jan. 4-6 and has a margin of sampling error of +/- 5%." (my bold italics).

Given the sampling error, a more accurate way to report this would be: "In the wake of Al Gore's decision not to seek the Democratic Party nominaion for president in 2002, the remaining hopefuls have statistically indistinguishable levels of support hovering in the high single digits."

I think John Edwards will be a serious contender for the nomination, and my guess is that Zogby believes this as well and is priming the pump. But the amount of breathless hyperbole in that copy -- given the reliability of the data -- is a bit nauseating. Maybe Zogby is plumping for Edwards, or maybe some public relations flack had way too many lattes before writing that.

UPDATE: An alert reader e-mails with the valid point that the sampling error decreases as the support numbers go to either extreme (single digits or above 80%). Still, I doubt the error figure would have declined to the point where Edwards' support is significantly greater than either Gephardt or Daschle.