About CFB Matrix


CFB Week #14 – The White Rabbit Beta Formula Tests

By  | 

Formula Beta Testing for College Football

Using numbers to find and edge ATS, Over, Under and SU


INFO: For those of you new to this page and other beta testing formula pages, I have simply decided to put some testing online for public consumption.   I am always searching for a story within the number and this testing is trying to find an edge in ATS, Totals and SU plays against the CFB Lines.  Remember this is just a test until we show a formula is a boom or bust.

UPDATE:  Another winner week.  This is starting to roll.   13-6 on spreads +/- 7.5 points and 6.0 went 4-4.  After three weeks of -2, -1 and -2 games under .500 the formulas take it all back .   The lines with 7.5 points or less continue a positive run following a strong 10-5 in week #12 and 9-7 in week #13.  After 9 weeks into the testing and formula 4.0 rises back to to 39 games above .500 ATS on spread under 8 points and 6.0 stays in the money.  Couple that with the NFL White Rabbit going 22-3-1 the last 2 weeks ATS and this is on a heater!

NOTES: In the previous weeks I have been putting up huge spreadsheets on the Beta page.  No more.  At this point 3.0, 4.0, 5.0 and 6.1 formulas have separated themselves into clear favorites and usefulness.  Here is how I will note them for each game going forward:

  • 3.0 is best used only when the away team has a better CFBMatrix talent ranking.  I lean towards it when both teams have decent Coach Effect numbers.
  • 4.0 Nothing trumps this formula to date.  She is the leader of formulas and the first one I look at for guidance.   After 8 weeks of testing, 4.0 is 29 games over .500 ATS.  Nuf sed.
  • 5.0  and 5.1 is a moving average.  I hope/expect this formula to gain strength in the latter half of the season as it phases out the OOC games and focuses solely on conference play.  My hope is that by November, this is the dominate formula.
  • 6.0 is not a formula but the manifestation of using the above formula in conjunction with the CFBMatrix team data base.  In the last 4 weeks the ‘special’ 6.0 teams are 20-9 ATS.    They are posted below as I find them.
  • TOTALS:  I think that totals mostly suck.  Probably because I cannot find a profitable pattern or formula.  It seems like it is 50/50 most of the time.  However there are clear times in which all formulas lean one way by over a touchdown.  I will post those leans too.

Week #14 – The Games

Under +/- 7.5 points

  • Texas -4.5
  • MSU +3.5 [OM if it drops to -2.5]
  • Nebraska -2.5
  • Akron/Toledo Pass – YTD model puts line at 7.35 in favor of Toledo
  • Marshall -3.5 [take ECU if you can get +4 or better]
  • PITT +2.5 (FWIW the modeling is VERY STRONG in favor of PITT covering]  #UpsetAlert
  • Duke +5.5
  • BC -2
  • NC State +2.5
  • Iowa State +8
  • Illinois +3.5
  • Rutgers in chalk but Flood is garbage.  Model says UConn cover but they are awful too.  Passing this game for the record
  • Kentucky +4 and Upset Alert?  That doesn’t look right
  • Missouri -4.5
  • Georgia Tech +3.5
  • Bowling Green -2
  • UNLV +3.5 [unlv much better in last 4 games]
  • Western Kentucky -6
  • Clemson +5 [spreads for each model .89, 4.47, 4.63 and 4.72 in favor of USC]
  • USC -3.5 [vs UCLA]

Over +/- 8 points

  • Michigan State -14.5
  • TCU +12

3.0, 4.0, 5.0 and 7.0 results and my opinions



Texas Tech at Texas -4.5

  • Score: 31.4-25.1 Texas (YTD)  44.1-25.4 Texas (last 4) 37.9-28.7 Texas (Conf play)
  • ATS:  All 3 major models favor Texas to COVER – strong
  • Total: Lean is very slightly OVER 66.5

Ole Miss at Miss State +3.5

  • Score: 27.6-26.9 MSU (YTD)  33.5-26.3 OM (last 4) 30.5-28.7 OM (Conf play)
  • ATS:  2 of 3 take Miss State to cover, 1 for MSU to win and game is on #UpsetAlert
  • Total: Lean is very slightly Under 53.5


Iowa at Nebraska

  • Score: 26.7-20.8 Neb (4.0) 27.8-21.3 Neb  (5.0 – conf games) 23.6-21.3 Neb(last 4 games)
  • ATS: 2 of 3 formulas say Nebraska to win and cover -2.5
  • Total: No lean- Range is on both sides of the line

Leave a Reply

Your email address will not be published. Required fields are marked *