How I pick a team:
-I would look through my spread sheet and pick out teams that were hitting, on a regular basis, one or both 1Q/3Q.
-I also had a list of fade teams and bet on teams.  Fade teams would be teams on the bottom of the rankings or consistently losing the 1Q/3Q..  Teams I would play would be teams on top of the rankings. 
-Other parameters would be the spread had to be 7 or less, because I saw a lot of losses for both dogs and faves for spreads 7.5 and up.

Changes in the system:
-Airmail26-A poster in my thread, hit 10-0-1 one day betting home teams only on the 1Q/2half.  I've looked at both the home fave and home dog angle in the past but back testing didn't do too well based on the parameters set above.  So I decided to revisit it once again and instead of putting any kind of parameters, I just looked at each home team regardless of dog or fave or spreads.  From the beginning of 09 season to Nov 11, 2009, the home team hit at about 79%.  In my thread, I stated that it was up 100 units, but I did the math wrong.  I counted the loss on the 3q but forgot to count the loss on the 1Q.  With that info, it wouldn't make a lot of money...
-Unless you use a martingale system to chase.  But if you're hitting almost 8 out of 10 teams, then you will win the chase.  But instead of doubling the 3Q, I think it would be profitable to spread the loss of a couple of games.