John Lott claims:
The final numbers will be out in March, but the initial information makes it look as if 2007 will be the coldest in a decade. So much for the claim that “the probability that 2007 would be the hottest year as 60 percent.”
NASA has 2007 tied for second, but he’s referring to the RSS version of lower troposphere temperatures, which have 2007, far from being the coldest in a decade, in fifth place:

Looks like someoen got caught by an error in the satellite data.