well, it's -3 now anyway. that said, it's (IMO) a losing proposition that I outlined a few days ago in one of the threads. paying double or triple juice catches up to you. every time you lose (which is 40%+ of the time) you pay extra juice. every time you win, you make nothing extra. you are betting (a lot) that it's going to hit exactly on one number. the risk/reward is not in balance.
this week, two games that closed at 2.5, 3 or 3.5 in the NFL:
NE -3 (-120) lost outright, costing the people that bought it down to -2.5 40%-45% juice instead of 20%.
IND -3.5 won by a few touchdowns.
look at something greater than a small sample size, and you see the long-term cost associated with this. penny wise and pound foolish.