I will give further explanation below, but based on a spreadsheet that I have been working on and improving over the last couple of seasons, and not looking at external factors like injuries, weather, etc., here are my spreadsheet picks for this week:
Falcons -3.5
Broncos -4
Bills +3.5
Vikings -3.5
Jaguars +1
Bears +3
Jets -6
Cardinals -3
Titans -2.5
Panthers +7
Packers -10
Ravens -2.5
Patriots -7.5
Eagles -4
First, let me say that I am not advocating placing these bets, and please keep in mind this is all based on stats that do not factor weather or injury. This has become a little hobby of mine, and I thought I'd put it out here. Since I don't place any money on it, there is nothing at risk unless I share it with others. Also, it means nothing if I come back after the games are played.
A couple of years ago I started playing with stats in Excel for college and pro football trying to see if I could come up with meaningful rankings, a la Sagarin. College was way too hard because of the number of teams, so I switched to NFL. Obviously the NFL doesn't use rankings like college, so I wanted to see if I could replicate spread predictions, and it started getting to be fairly close. Then I had the bright idea of tracking my predictions vs. Vegas lines to see which side I come up on.
Last year was the first full season I got done, although some of it was retroactive, and I wasn't sure if the lines I was finding were accurate. It still came up over 50%. You need 52.4% to make a profit, albeit a small one at that number. So far this year it's at 57.6%. As the season goes along, if it works correctly, the fluctuations from week to week should get smaller and more accurate because there are more data points. It would take a while with some visual aids to explain the exact formulas I came up with, and there are probably a lot of similar ones, but I'm happy with it so far, and have more ideas for improvements going forward. Now that I am pulling the actual lines (every Thursday at 9:00 AM from Vegas Insider, using MGM spreads), any changes I make can be applied to past seasons to test their accuracy.
Anyway, I hope this works. If anyone reading this has any similar experience and wants to collaborate on future improvements, hit me up.
Falcons -3.5
Broncos -4
Bills +3.5
Vikings -3.5
Jaguars +1
Bears +3
Jets -6
Cardinals -3
Titans -2.5
Panthers +7
Packers -10
Ravens -2.5
Patriots -7.5
Eagles -4
First, let me say that I am not advocating placing these bets, and please keep in mind this is all based on stats that do not factor weather or injury. This has become a little hobby of mine, and I thought I'd put it out here. Since I don't place any money on it, there is nothing at risk unless I share it with others. Also, it means nothing if I come back after the games are played.
A couple of years ago I started playing with stats in Excel for college and pro football trying to see if I could come up with meaningful rankings, a la Sagarin. College was way too hard because of the number of teams, so I switched to NFL. Obviously the NFL doesn't use rankings like college, so I wanted to see if I could replicate spread predictions, and it started getting to be fairly close. Then I had the bright idea of tracking my predictions vs. Vegas lines to see which side I come up on.
Last year was the first full season I got done, although some of it was retroactive, and I wasn't sure if the lines I was finding were accurate. It still came up over 50%. You need 52.4% to make a profit, albeit a small one at that number. So far this year it's at 57.6%. As the season goes along, if it works correctly, the fluctuations from week to week should get smaller and more accurate because there are more data points. It would take a while with some visual aids to explain the exact formulas I came up with, and there are probably a lot of similar ones, but I'm happy with it so far, and have more ideas for improvements going forward. Now that I am pulling the actual lines (every Thursday at 9:00 AM from Vegas Insider, using MGM spreads), any changes I make can be applied to past seasons to test their accuracy.
Anyway, I hope this works. If anyone reading this has any similar experience and wants to collaborate on future improvements, hit me up.
