Just something I was messing around with 2 nights ago. It accurately predicted the score of the PHX game yesterday and on both occasions, it was on the winning side of the total. Here is the formula:
A + B + C = TOTAL
A= (FGM per game - 3PM per game)/Points Per Possession*2
B= 3PM per game*3
C= FTM per game
In order to calculate Points Per Possession:
Points per game/(FGA per game + Turnovers per game)
Now whether or not, this formula makes complete sense or any sense at all, I just want to see where it goes. I know there is no account for defense, or how many games played or any of that, it's purely offensive stats. If it works, great, if not, no big deal, no money lost. It's just experimental.
Using this formula, yesterday it predicted a total of 232 in the Phx game and 208 in the Minnesota game. Thus resulting in the picks Phx/Atl OVER and Min/Hou UNDER. Well, tonight I used this formula for every game and these are the totals and their resulting plays for tonight: