A few weeks ago I posted some NFL odds:
Included were estimated point ranges for NFL teams. In my opinion this was a great tool to help predict NFL wins for the season.
So I ran a regression to figure out the formula between points scored and number of wins for the last three seasons. The formula was about 60% predictive. Ideally you want this as high as possible, but that would include adding many more stats and estimating quite a few categories will introduce their own errors, so I went with this for now.
Pittsburgh and NE weren't evaluated since their bets were not listed in my book.
So using the under # points and over # points listed in the above thread for each team, plugging it into my regression formula, I came up with an over # and and under #, which I also have the bookies odds for as well
If a team was listed at say 7 wins, and my estimations came out at 6 and 8, I did not evaluate the bet. Those teams are: Atl, Bal, Buf, Jax, LA, Mia, NO, NYJ, Was
Included were estimated point ranges for NFL teams. In my opinion this was a great tool to help predict NFL wins for the season.
So I ran a regression to figure out the formula between points scored and number of wins for the last three seasons. The formula was about 60% predictive. Ideally you want this as high as possible, but that would include adding many more stats and estimating quite a few categories will introduce their own errors, so I went with this for now.
Pittsburgh and NE weren't evaluated since their bets were not listed in my book.
So using the under # points and over # points listed in the above thread for each team, plugging it into my regression formula, I came up with an over # and and under #, which I also have the bookies odds for as well
If a team was listed at say 7 wins, and my estimations came out at 6 and 8, I did not evaluate the bet. Those teams are: Atl, Bal, Buf, Jax, LA, Mia, NO, NYJ, Was