HomeCOLLEGE FOOTBALLCOLLEGE BASKETBALLPRO FOOTBALLOTHER SPORTSABOUT USCONTACT USDonate

Update (Wednesday, Sep17, 2014 1:21pm): The models are ready and there are 16 picks for this weekend's games; 15 "3-Star" games and 1 "2-Star"  Historically, Week 4 games have won at 70% (49-21-1). 
 

How the system works:
Each game is run through my models to see what the projected margin of victory will be.  That projected outcome is compared to the Vegas pointspread to find which games are the best betting opportunities (i.e. the games where the pointspread is farthest off from what my model says it shoule be).  If there is a big enough difference between what i think the margin of victory will be and what the pointspread is, I flag it as a pick (a game on which you should bet).
 
Why there are 4 different models?
There was never a plan to construct four separate models.  In 2006 I built a model and was very succesful with it.
   
After the season was over I found some things in the model that I felt could be improved.  So, I built a new (what I thought would be a "better") model for the 2007 season.  That year was a good one as well, though it identified fewer picks.
 
I decided to use both models in 2008 to compare them head-to-head.  The results were about the same.  Both won approximately the same percent of games, though the 2006 Model flagged more games as picks.
  
I decided I could do better yet, and was also interested in building what would - on my end - be a less cumbersome model to operate.  So I built a new model for the 2009 season, and ran it against both of the older models.  I was pleased with the results of the 2009 Model, not only because of the winning percentage, but also because of the high number of games it flagged as picks.
  
Finally, in 2011 I came up with a model that was far simpler to operate than any of my older models.  I ran all four models in 2011.  My hope was that the 2011 Model would be the best one and I could then discontinue the use of the other three, as running four models is quite time consuming. 
Thus far, the 2011 Model has been (ever so slightly) the best one.  But for now, I have decided to continue to run the older models as well.  I've noticed that which particular combination of models identifies a game as a pick can be used to further identify likely winners. 
Based on which models identify each game as a pick determines the likelihood of that pick being a winner, and what Star Rating the game will be given (1-star, 2-star, 3-star or 4-star).


 

 



See link below for the latest predictions: 

Picks Here

Below are the results of all past predictions:

  2006 Model2007 Model2009 Model2011 Model 1-star2-star3-star4-starAll Games
Yr.Wk.WLTPct.WLTPct.WLTPct.WLTPct. WLTPct.WLTPct.WLTPct.WLTPct.WLTPct.
200640000%0000%0000%0000% 0000%0000%0000%0000%0000%
200650000%0000%0000%0000% 0000%0000%0000%0000%0000%
20066100100%0000%0000%0000% 0000%0000%0000%100100%100100%
2006731075%0000%0000%0000% 0000%0000%0000%31075%31075%
200680000%0000%0000%0000% 0000%0000%0000%0000%0000%
20069200100%0000%0000%0000% 0000%0000%0000%200100%200100%
2006100000%0000%0000%0000% 0000%0000%0000%0000%0000%
2006110000%0000%0000%0000% 0000%0000%0000%0000%0000%
2006120000%0000%0000%0000% 0000%0000%0000%0000%0000%
2006130100%0000%0000%0000% 0000%0000%0000%0100%0100%
2006140100%0000%0000%0000% 0000%0000%0000%0100%0100%
2006150000%0000%0000%0000% 0000%0000%0000%0000%0000%
2006160000%0000%0000%0000% 0000%0000%0000%0000%0000%
2006171002100%0000%0000%0000% 0000%0000%0000%1002100%1002100%
2006 Total163284%000####000####000#### 000####000####000####163284%163284%
200740000%400100%0000%0000% 0000%0000%400100%0000%400100%
200750000%31075%0000%0000% 0000%0000%31075%0000%31075%
200760000%0100%0000%0000% 0000%0000%0100%0000%0100%
200770000%0100%0000%0000% 0000%0000%0100%0000%0100%
200780000%0100%0000%0000% 0000%0000%0100%0000%0100%
200790000%0000%0000%0000% 0000%0000%0000%0000%0000%
2007100000%0000%0000%0000% 0000%0000%0000%0000%0000%
2007110000%0000%0000%0000% 0000%0000%0000%0000%0000%
2007120000%0000%0000%0000% 0000%0000%0000%0000%0000%
2007130000%0000%0000%0000% 0000%0000%0000%0000%0000%
2007140000%41080%0000%0000% 0000%0000%41080%0000%41080%
2007150000%0000%0000%0000% 0000%0000%0000%0000%0000%
2007160000%0000%0000%0000% 0000%0000%0000%0000%0000%
2007170000%0000%0000%0000% 0000%0000%0000%0000%0000%
2007 Total000####115069%000####000#### 000####000####115069%000####115069%
200840000%31075%0000%0000% 0000%31075%0000%0000%31075%
2008542067%43057%0000%0000% 11050%12033%31075%0000%54056%
200860100%0000%0000%0000% 0100%0000%0000%0000%0100%
2008721067%0000%0000%0000% 21067%0000%0000%0000%21067%
200880000%100100%0000%0000% 0000%100100%0000%0000%100100%
20089100100%100100%0000%0000% 100100%100100%0000%0000%200100%
2008100000%0000%0000%0000% 0000%0000%0000%0000%0000%
2008110000%0000%0000%0000% 0000%0000%0000%0000%0000%
2008120100%0000%0000%0000% 0100%0000%0000%0000%0100%
200813100100%100100%0000%0000% 0000%0000%100100%0000%100100%
2008140000%0000%0000%0000% 0000%0000%0000%0000%0000%
2008150000%0000%0000%0000% 0000%0000%0000%0000%0000%
2008160000%0000%0000%0000% 0000%0000%0000%0000%0000%
2008170000%0120%0000%0000% 0000%0120%0000%0000%0120%
2008 Total85062%105267%000####000#### 44050%64260%41080%000####149261%
200940000%200100%0000%0000% 200100%0000%0000%0000%200100%
2009521067%0100%43057%0000% 0100%42067%11050%0000%54056%
2009662075%101100%21067%0000% 0010%33050%300100%0000%63167%
2009731175%200100%71088%0000% 100100%72178%200100%0000%102183%
20098100100%0100%22050%0000% 0100%32060%0000%0000%33050%
2009911050%0000%32060%0000% 0000%43057%0000%0000%43057%
2009100000%0000%31075%0000% 0000%31075%0000%0000%31075%
2009110000%0000%101100%0000% 0000%101100%0000%0000%101100%
2009120100%0000%53063%0000% 0000%54056%0000%0000%54056%
200913100100%0000%0000%0000% 0000%100100%0000%0000%100100%
2009140000%0000%100100%0000% 0000%100100%0000%0000%100100%
2009150000%0000%0000%0000% 0000%0000%0000%0000%0000%
2009160000%0000%100100%0000% 0000%100100%0000%0000%100100%
20091764060%0100%200100%0000% 0000%63067%11050%0000%74064%
2009 Total2010167%53163%3113170%000#### 32160%3920266%72078%000####4924367%
201040000%33050%0000%0000% 33050%0000%0000%0000%33050%
2010542067%22050%71088%0000% 22050%82080%100100%0000%114073%
2010621067%100100%55050%0000% 100100%34043%21067%0000%65055%
20107100100%100100%62075%0000% 100100%62075%0000%0000%72078%
201080100%11050%42067%0000% 11050%41080%0100%0000%53063%
20109100100%0000%22050%0000% 0000%32060%0000%0000%32060%
2010100000%0100%21067%0000% 0100%21067%0000%0000%22050%
2010110000%0000%11050%0000% 0000%11050%0000%0000%11050%
2010120000%0000%0200%0000% 0000%0200%0000%0000%0200%
20101311050%100100%100100%0000% 0000%11050%100100%0000%21067%
20101411050%0000%11050%0000% 0000%0200%100100%0000%12033%
2010150000%0000%0000%0000% 0000%0000%0000%0000%0000%
2010160000%0000%0000%0000% 0000%0000%0000%0000%0000%
20101794169%52071%21167%0000% 0000%21067%73170%0000%94169%
2010 Total1910166%149061%3118163%000#### 87053%3019061%125171%000####5031162%
201140000%43157%0000%94169% 0000%31075%73070%0010%104171%
2011531075%11050%84067%186075% 31075%100100%145074%100100%196076%
2011632060%300100%0600%12033% 15017%11050%11050%100100%47036%
2011711050%0000%32060%22050% 23040%100100%22050%0000%55050%
2011811050%11050%21067%100100% 12033%21067%100100%0000%43057%
2011911050%100100%23040%0000% 12033%21067%0000%0000%33050%
2011100000%0100%500100%0000% 500100%0100%0000%0000%51083%
201111100100%200100%0000%0000% 0000%100100%0000%100100%200100%
2011120000%0000%200100%0000% 200100%0000%0000%0000%200100%
20111312033%0000%200100%0000% 12033%100100%0000%0000%22050%
201114300100%0000%0100%0000% 31075%0000%0000%0000%31075%
2011150000%0000%0000%0000% 0000%0000%0000%0000%0000%
201116100100%0000%200100%0000% 300100%0000%0000%0000%300100%
20111744050%0000%0000%0000% 44050%0000%0000%0000%44050%
2011 Total1912061%126167%2617060%3114169% 2620057%125071%2511069%301100%6636165%
201240000%64060%0000%81089% 0000%33050%41080%300100%104071%
2012561086%400100%52071%74064% 52071%200100%63067%200100%155075%
2012655050%42067%42067%32160% 13025%21067%31175%32060%97156%
2012742067%11050%0500%200100% 46040%0000%200100%0000%66050%
2012831075%100100%22050%0000% 43057%0000%0000%100100%53063%
2012931075%200100%42067%0000% 32060%300100%0000%0000%62075%
201210100100%100100%11050%0000% 11050%0000%0000%100100%21067%
201211100100%12033%22150%0000% 22150%0200%0000%100100%34143%
201212200100%21067%0000%0000% 100100%11050%0000%100100%31075%
2012130100%0100%200100%0000% 21067%0100%0000%0000%22050%
2012140200%0000%0000%0000% 0200%0000%0000%0000%0200%
201215100100%0000%0000%0000% 100100%0000%0000%0000%100100%
201216200100%0000%100100%0000% 100100%100100%0000%0000%200100%
20121787153%400100%0200%0000% 54156%0200%0000%300100%86157%
2012 Total3620164%2611070%2118154%207174% 3026254%1210055%155175%152088%7243363%
201340000%115169%0000%138162% 0000%63067%86057%301100%179165%
2013543057%21067%45044%55050% 34043%0200%45044%200100%911045%
2013633050%31075%11050%51083% 44050%31075%51083%0000%126067%
2013712033%0100%31075%0000% 22050%100100%0000%0100%33050%
20138200100%0000%42067%0000% 42067%100100%0000%0000%52071%
20139100100%0100%0410%100100% 14120%0100%100100%0000%25129%
2013100200%0100%23040%0000% 23040%0000%0000%0100%24033%
201311100100%200100%22050%0000% 21067%000