I went back over this season (quickly, but I think accurately), and was actually surprised to discover that buying the .5 isn't as bad as I thought it would be. If my numbers are off, I'm glad to hear about any corrections. Usually the lines in my spreadsheet are what come out of WSEX on Wednesday.
First of all, I found 6 pushes, and 9 games that fell within a half point of the number. 192 games have been played so far.
Consider 3 people who are playing $110 per game.
Let's say Average Joe is 93-93-6 and so is $930 in the hole paying -110 vig. If he bought all his games with a .5, he'd, for starters, turn the pushes into wins and be 99-93, but with -120 vig, he'd be $1260 in the hole. However, you need to turn some of those losses (let's say 5 of 9) into pushes and that gets him to 99-88-5, or $660 in the hole. Hmmm.
Now, look at a winner. He's 100-86-6 and $540 to the good on the year. Turn the pushes into wins, and he's 106-86, or $280 to the good. Turn 5 losses into pushes, though, and he's 106-81-5, or $880 to the good. Hmmmm. Hmmmm.
Now, look at a loser. Let's say he's 86-100-6. He's down $2400 on the season. Turn the pushes into wins and he's 92-100, but $2800 in the hole. Turn 5 losses into pushes and now he's $2200 in the hole.
Conclusion: Well, surprisingly, it really doesn't seem as bad as I thought buying that half point and actually looks beneficial in some cases. You need to hit about 52.4% to break even with normal vig, and 54.5% to break even laying -120. The .5 certainly seems to be figuring into enough games this season to make it worth it. If you bought it off key number, only, these results might even improve. I'd be interested if anyone had statistics on how often the game fell within a .5 point of the number over many seasons, not just this one.
These numbers support what GM said about losses into pushes being much more valuable than pushes into wins, which is a good point.
Final note: With the 9 games that fell within a .5 point of the spread, I only gave half of them to the bettor. Why? Consider, say, CHI @ GB this week. All the GB bettors would have bought the half point down, making no difference, while the CHI bettors would have bet it up, turning the loss into the push. I assume our average bettor would be about 50/50 on these. Maybe he had Minny against NE last week (no diff) and CHI this week (loss -> push).
The other thing this tells me is how important it is to shop for lines. For instance, I had raiders -6.5 last night. If I still had money at another book, of course I would have gone with 6 if I could.