My own NCAAF system (part 2)

Figured I would just start a new thread rather than clutter the other one now that I have run some final filters to improve it.

System: After first few weeks of season, Bet against any team getting 20 or more points who's run defense averages over 5 yards per carry.

Filter 1: Cancel out any team that fits the system if their run offense averages less than 4 yards per carry.

Filter 2: Only play teams that fit the system if their run offense averages over 5 yards per carry.

Numbers:
System with no filters over last 4 years: 63% ATS (profit every season)

System after filter 1 over last 4 years: 69% ATS (was 69% - 70% every single season.  How is that for consistency?)

System after filter 2 over last 4 years: 78% ATS (yes, you read that right, 78% on a sample size of over 30 games)

note: Grab them as soon as the lines come out as these type of lines typically go up 1-3 points from the opening line over the course of the week.

Picks that meet the system this week:

No Filter: East Carolina -20

Filter 2: West Virginia -20

Great work Pistol Pete............
Honorable to mention to Kmang to for the suggestion..........

Would it be an idea to attach a unit structure to the strength of the plays with or without the filters?



Say maybe like

No filter  =  1 unit
Filter 1   =  2 units or so
Filter 2   =  5 units or so

Any variation of the sorts

Just a suggestion don't mind me, thanks guys.........

I will just track the record in here and let anyone who decides to follow apply whatever amount of units per filter they seem fit.

Also, on the picks this week, I was going by the early wagerline lines.  They could end up being non plays, seems like the few sportsbooks that have lines already posted are opening the West Virginia line at 18.  I also tested this theory/system against 15-20 point lines and the results were mediocre which is why I made the filter 20+.

Kiss - Then maybe we should wait to bet the line? Instead of getting it right away incase it falls?

I will wait until tomorrow and see what Vegas opens it up at.

Interested to see how this works. If you looked at Historical data then the lines you were looking at had to be closing lines right?

Which makes me think that anything over 20 before game time is a play?

Do you have data of let's say -25? or -30? I guess I'm wonder how far off of -20 this will work for?

Yeah, I used closing lines.  and I don't have the specifics of winning percentages for specific line ranges over 20 and it is not something I can just pull up quickly.  I saw nothing in the stats to indicate the system needed a ceiling filter. 

I will just post my research and leave it up to anyone who decides to follow to use their own judgement on questionable plays that sit around my 20+ point cutoff filter or plays pushed to one side of that cutoff point or the other due to line movement.  I personally plan on going by whatever the sportsbook I use personally has the line at on Monday night since that is when I like to make my wagers for college football games to get the best value on most wagers, if it is below 20 I plan on leaving it alone.  But 19 or 19.5 may be "close enough" for some people.  Your call.

My sportsbook has East Carolina opening at -18.5 and W. Virginia at -19, so I personally are going to void those two and not make any wagers this week, I will wait and see what comes up next week.

I only had access to closing lines so that is what those numbers reflect.  Going by the the opening line is just what I plan to do personally.  Like I mentioned, I started tracking the system for lines 15+ to start and games with 15-19.5 lines were cut off because they produced mediocre results.

Two plays this week:

Filter 2: California

Filter 2: Fresno St.

Also have one other system I came up with myself that kicks in after week 5 and is off to a 5-1 start.  South Carolina ATS is the only pick for that system this week.

Got Cal at -36 and Fresno St. at -25. 

S. Car at -12.5 for system 2.


Also, recommended 2-team parlay (based on the results I got from my system #2).  I play these any time I find two -13 to -17.5 home favorites who according to my program should be even larger favorites and I parlay them on the ML for about a 40% return of the risk. 

Alabama ML
S. Carolina ML

Pistol,  with you on Cal and Fresno.  What can you tell me about the system 2 that identified S. Carolina as a play?  I have seen Vandy as a play on a different thread in this forum (Kangy's NCAA JM thread had something about playing on Vandy this week). 

It's not your typical "trend" system.  I have this pretty complex program I use to come up with my own lines for the games that I have used for years.   Through trial and error I noticed which worked and which didn't.  The program seems to undervalue home field advantage.  So, I take home teams any time they should be giving or getting a certain amount of points more than they are (according to the program) barring there is no key injury to account for the line being off.  Taking road teams with lines that were off always seemed to just eventually end up 50/50 so I have limited it to just home teams now with good results.

Here are the results so far:

last week:
Ohio -12.5 / program had them at -21 / Ohio won by 21 (W)
Indiana +1.5 / program had them at -7 / Indiana won by 13 (W)
BC -2.5 / program had them at -11 / BC won by 32 (W)

two weeks ago:
Minnesota -3 / program had them at -13 / Minn won by 15 (W)
Wyoming -10 / program had them at -18 / Wyo won by 14 (W)
LSU +8.5 / program had them at PK / LSU lost by 10 (L)

Program has the S. Car line way off this week and has S. Carolina at -25.

System could have easily been 2-0 with Cal being up 35-3 in the first half....its like they stopped playing and at the end of the game they were driving but used up the rest of the clock.

Lets sweep next week!!

 

Quick question KAP,

 

where do you check your stats to find your plays? 

Thanks

Nevermind, I have a great way here at covers!

Quite a few games fit the system this week:

No Filter: Notre Dame -29
No Filter: Arkansas -36.5

Filter 1: TCU -33
Filter 1: Ohio St. -38.5

Filter 2: Boise St. -34.5

System 2 also has a lot of plays this week:
Iowa -17
Idaho -3
Baylor +12.5
N. Illinois -3
Oregon +3.5
*wouldn't recommend taking any of these if the line is a point or more off from what I got them at due to line movement during the week.  Well, except Iowa, that would be a play as long as the line stays under 20.

Recommended 2-team parlays:

Iowa ML
SD State ML

Utah ML
Fresno St. ML

n.illinois is -12 you should recheck that game and

My bad, that must of been a typo.  I got N. ILL at -10.   Technically, that game only fits the system if you can get it at -10 or better.  Line has moved quite a bit on that one already.

Dude how many wins have you had with this system so far...??

BOL