Sent as an email rather than as a comment was this idea from Phil:
BLUnders seems great... what about SLOvers?
Hi CassiniLove the input. An exchange of ideas is what this blog should really be all about. Limiting selections to games where the spread is zero means a very small sample size, somewhere around 24 bets a season.
Love the blog. Thought I'd just raise what (to me) is an obvious question. If BLUnders works, then assuming that bookmakers correctly price 50% of games to hit the over and 50% to hit the under, does betting overs when teams have no spread (pk) make money?
Above are the results from the 2010 season on, one good season for Phil's SLOvers, but no evidence that there is any edge with this strategy.
If we expand the range of selections to include matches where the spread ranges from 0 to 2.5 points, the numbers for the last five years look like this:
Again, one good season, but the Over / Under ratio over these five full seasons plus 2015 is pretty close to 50:50 and we are looking for trends, not single season blips.