To Any of you guys with your own computer rankings.....

[Lehigh'98]

Do you use your systems to predict FBS, FCS or NFL game scores? Do you win more than u lose if you do against the spread? I have a simple system based on moving averages for various stats that I'm testing in the NFL this year, hoping to hit above 58%, but would be interested in seeing a more advanced system in play.

[Wallace]

Kenneth Massey maintains a website of many computer rankings together if you want to see how others do it: http://www.masseyratings.com/

[gumby013]

I don't use mine for betting. It does predict FBS and FCS games though. I'm going to West Point tomorrow for Eastern Michigan vs Army, and it's predicting an Army win 76.3% of the time.

[Lehigh'98]

Thanks Gumby. The pointspread gives me a pretty good idea on the % chance of winning. Unless there is a big discrepancy they should be very correlated. I am looking for predicted scores for the past couple years so I can compare the predicted pointspread from the program vs. the actual outcome of the game. I'd be interested to see if they are much better than 50/50 vs the spread.

[TILIS-BisonFan]

Thanks Gumby. The pointspread gives me a pretty good idea on the % chance of winning. Unless there is a big discrepancy they should be very correlated. I am looking for predicted scores for the past couple years so I can compare the predicted pointspread from the program vs. the actual outcome of the game. I'd be interested to see if they are much better than 50/50 vs the spread.

http://espn.go.com/blog/playbook/dollars/post/_/id/2935/meet-the-worlds-top-nba-gambler

interesting read and Im fairly certain no "simple" formula will show long term profit. Lines are shifted by professional gamblers and also not set by simple statisticians. Professional gamblers are future line makers

[gumby013]

Thanks Gumby. The pointspread gives me a pretty good idea on the % chance of winning. Unless there is a big discrepancy they should be very correlated. I am looking for predicted scores for the past couple years so I can compare the predicted pointspread from the program vs. the actual outcome of the game. I'd be interested to see if they are much better than 50/50 vs the spread.

I haven't looked at the predictions vs the spread yet. I hadn't done an official release until this past week though. Even now I'm not thrilled with the limited links between teams. The Ivy League isn't helping things by starting 3 weeks later than everyone else. The rankings and predictions will continue to improve as more games are played.

[Lehigh'98]

For the NFL, I'm using data from 2003-2013. Grouping games into 4 diff categories depending on point spread and combining some key stats to create a running MAvg for the prior 3 home or away games. Some stats show a 60% or above win vs spread depending in which group they fall. Nothing really works for the sample data as a whole ( no groups). I'd really like to compare to a computer system though. Thanks for links, I'll see if I can do anything with them.