Quote:
Originally Posted by MyTurn2Raise
who said I'm treating the team outcomes as independent?
pretty sure I haven't disclosed how I'm doing it
OK now we are getting somewhere. I'll take that as an acknowledgment you agree the Team Totals should not be treated independently when assigning probabilities to your scores. Well, one of the things that they depend on is
who is the damn underdog. If Team A is a big fave over Team B and you give Team B a Team Total of 21, Team A's push rate of 20 goes down simply because they are the big pregame favourite. Or put it this way, Team A's push rate of 20 would be higher if they were favoured by less. You have to apply this variable in such a way that holds true to the intuitively obvious notion that an underdog favoured by x can never be a good bet at +100 over the fave winning by x.
Likewise you can't stare at something like a DAL by 4 : DAL by 5 ratio of 2.5, see the Gaints' counterpart at 1.5 and not get worried about that either. There's just no reasonable explanation for push rates to fluctuate like that. I'll bet you anything if you start tweaking the market lines you are dealing with up or down a couple of points you will not be comfortable with how much you see your push rates start bouncing up and down.
You can be the world's best stathead but you're not going to get anywhere if you can't interpret your results properly. Here you have obvious red lights screaming at you that there's something wrong with your methodology. You have to find out what it is and fix it.