There seems to be more than a few handicappers lately suggesting that we buy 1/2 points on certain lines - most recently DeMarco adamant that we buy the half point with Minnesota regardless if the line is 3.5, 4 or 4.5.
Unfortunately this is not sound logic - not specific to this game but taken on the whole.
While I do see merit in buying from 3.5 to 3 (usually -135), it's still a very expensive proposition. Paying triple juice, betting that it will be exactly a 3-point game doesn't compensate you from a risk-reward perspective. Yes, you get "saved" occasionally, but 40%-50% of the time (when you lose) you are paying a fortune. When you win, it's irrelevant, but it catches up to you very quickly when you're on the wrong side.
Let's assume that you are extremely lucky and win 60% of the time:
100 games with a line of 3.5
60 wins = +6000
40 losses = -4400
Net = +1600
100 games with a line of 3.5, buying a half point
60 wins = +6000
40 losses = -5400
Net = +600
You need eight games (of 40) to be exactly 3 points to make profit over holding at 3.5 - even with seven pushes, you are still behind the +1600 from earlier.
60 wins = +6000
33 losses = -4455
7 pushes = 0
Net = +1545
It's not nearly that frequent.
Sure as shit, it will be a 3-point game tonight, but the odds are against it. One game yesterday was -3.5, and Tennessee (the favored team) lost by 20 - costing the 1/2 point buyers $135 on their $100 wagers.