I love it when the pundits are wrong:

Several reports these last few days have noted that Howard Dean had bagged about 8,000 absentee votes in New Mexico prior to his post-Iowa tailspin. This in itself should have made him competitive in New Mexico, since, as most of these reports went on to suggest, only about 50,000 people were expected to vote in today's state-wide caucuses. In fact, combine the 15 percent of the likely vote Dean's cache of absentee ballots gives him with the 15 percent support he garnered in the final New Mexico poll, and he's basically in the same place, 30 percent, John Kerry is (at least according to the same poll).

But now come reports that there's a major snow storm headed toward New Mexico (and the Albuquerque/Santa Fe population hub in particular) this evening. If the storm pushes turnout significantly lower than the expected 50,000, it doesn't seem like a stretch to think Dean could actually carry the state by a relatively comfortable margin.

(Never mind that his arithmetic is flat wrong, that Dean's alleged 8,000 absentee votes plus 15 percent of the number Scheiber uses above gives him must 23.5 percent of the vote. That really doesn't matter. He could have just waited a couple of hours to see what actually happened, rather than predicting the future, badly. What's the rush?)