Your ATS ranking methodology seems to have an underlying assumption that nothing changes during the year to make a team have a different probability of beating the spread.
1. A team like Alabama almost never covers the spread when it is playing a cream puff team from a non-power-5 conference. The public expects Alabama to beat the creampuff by 40 pts -but 12 out of the last 12 times it has not happened. You can make a lot of money if you understand that.
Conversely, Alabama often beats the spread against an SEC team that is good but not great: the players get up for the game, or whatever.
Your ATS analysis will not pick up that pattern in Alabama's ATS performance.
2. The spread often underestimates the impact of injuries to players that occur in the last week or two -or the impact from the return of injured players.
3. The spread often overreacts to the previous week's performance -particularly when a team is surprisingly awful or surprisingly dominant. You might analyze how well teams do against the spread a week after they beat the spread by 14 points or a week after they failed to cover the spread by 14 points.
4. A new quarterback can improve greatly during a season (ditto for an offensive line or a defense with a new coordinator) -causing a team to underperform ATS early in the year and beat the spread later in the year.
I'm not saying that your ATS analysis is worthless -it will probably pick up some interesting trends. But "asking the right questions" is important to any mathematical analysis, and its not clear to me that you are asking the right questions.
Can you taste how salty Axel is that I don't kiss his ass all the time??? ;-)Quote: AxelWolf
...Don't let it go to your head, Romes kisses a lot of ass. (-;
JoelDeze, thanks for your reply about your methodology - it does make me place more weight on your recommendations.
Many years ago, I used to calculate power ratings for CFB and NFL teams -but I've decided that ESPN's FPI is about as good as almost anyone can do. And I realize that the sportsbooks use a very sophisticated proprietary model to calculate power ratings as a basis for setting their opening lines.
So, instead of reinventing the wheel on power ratings, I now try to pay attention to factors that I categorize as "inefficiencies in the line" to see where I might get an edge. But I'm always learning and anxious to learn more. I'll continue to read your predictions. Good luck with this.
Any chance you could do NFL predictions? I know 'a little' about the NFL and nothing about college and would like to see if the very very little I know/research is anywhere along the same thought paths that you take.
While some controversy, I again still just want to say, so long as you get the picks in before, I enjoy reading about your thoughts and seeing how you do as more data is collected - so thanks for that yet again!
Can you taste how salty Axel is that I don't kiss his ass all the time??? ;-)