I just went though and got my site ready for the new season. Like always week 1 is just based off last season statistics and usually not enough data is gathered till after week 4 to make accurate predictions. Feel free give me any feedback if you want as well I will bump again week 5 when I have my Bet Busters up http://www.collegefootballcalc.com
Curious about your calculation process. Is it power rankings based or based on PF/PA percentages, etc..
It's gotten more in depth every year... but basically i take passing yardage, rushing yardage, turnover ratio and turn that into an equation for scoring, then I take a defenses passing yards given, rushing yards given, and turnovers forced into another equation. I then weight each team based on a bunch of factors and then compare the offensive and defensive results. Then I "average" them together for lack of better word and I get an estimated score, yardage, and turnover % Then I take the top 10 score differences between my guesses and Vegas and I bet on a couple of those and I have had pretty good success beating Vegas that way. Last year i also started tracking how a team performs against the spread, so I use that to pick 1 game out of the 10 biggest differences for my most likely bet buster
Do you have the stats for how it performs for favorites vs underdogs? Or games where you have vegas with the wrong team favored vs games where your line is different but the same team is favored? Or where your line is within 3 points...6 points...double digits? Would be interesting to see if there's any trends there.
It's still a young project and something I develop on the side. Right now I just save the results of my top 10 differences with Vegas and my overall % for the week for win/loss and vs Vegas. Improvements I want to add this year is to start saving more data I produce so I can generate some charts and graphs. Bitches love visual data.
It sounds similair to something I created a good while ago. Was pretty good predicting lines after about 6 weeks after enough stats were in. Was money on over unders when the difference was more than 10 points. Would be glad to share it. Never factored in SOS. That was hard to figure.
I remember this from last year. So...what are your % like for you top 10 differences w/ vegas & your overall?
Week 1 uses the final stats from last year, and week 2 only uses week 1 statistics The calculator isn't using enough data till about week 4 but last week I had no top 10 (for the reasons listed above) This week over Vegas we correctly picked the winner 46.67% of the time This week we correctly picked the winner 93.33% of the time!
LOL, i would say wait till next week when the stats are starting to even out. But i really think UCF playing FIU is just gonna make our stats look even better. UCF is having a statistically freakish year
My betting lines are finally posted Honestly this weeks 10 bets look terrible... Probably should have waited for another weeks worth of statistics