as djv sez, and zoomer mentions, the overall betting trend of the home dog in NFL is "a starting point."
I've met guys in sports books who fine-grained this trend more, by especially taking home dogs playing in a second straight home game, or vs. a divisional foe.
but they never could give me stats to demonstrate if this refinement of data was really as worthwhile as it sounds.
I am not enough of a trend bettor to chase down the stats, and neither do I know where to look....
one fellow I met drilled the data down to just the first 3 weeks for the season, and did give me the stats for a few years:
NFL home dogs:
first 3 weeks, 2012 19-6 ATS
first 3 weeks, 2011 13-3 ATS
first 3 weeks, 2010 12-7 ATS
But! This Season to Date, dogs have gone 4-12-0, home dogs 1-4-0.
Anyway, for background, here is
link to Leavitt's article on betting home underdogs, from 2009.
and here is a paper (warning,
it is a paper in pdf format) that regressed the bet outcome to "large closing lines", and uncovered a winning percentage of 60% (for the years 2000-2008) :SIB
someone should gather the data, update the predictive analytics for that trend, and post it here!
opcorn2
Link here to a paper that found betting home dogs late in the season is a profitable way to go