I'm in the media room at the Dean campaign-night event here at Southern New Hampshire University.
About an hour ago, a little birdie inside the Dean campaign informed me that day-long ID calls of Dean two's and three's (i.e., leaners and uncommitteds), are breaking by a ratio of 2.5-to-1 for Dean. (For the math-challenged, that translates to 72 percent.) Now, if true, that's good news for Dean. But keep in mind that these sorts of calls no doubt include systematic biases because (a) they are being dialed by volunteers; (b) voters at this point will probably say almost anything to get callers off the line; and (c) who knows whether that ratio is accurate in the first place? (However, the degree of specificity, coupled with the trustworthiness of this person, leads me to trust it, biases "a" and "b" still applicable.)
Cannot get my contact inside the Kerry camp on the phone, but I'm trying....stay with me.