cyptomcat's Season Thread

[cyptomcat]

I want to make some public picks from an algorithm that has done well so far this season. Let's see how next several weeks go.

USC -21.5
Louisville +14
Oklahoma -13
Northwestern +17.5
Wisconsin -25.5
Arkansas -4.5

With 100 units, let's put 16.66 on each.

[DerekMadden]

I like it..

I have all the same picks

[cyptomcat]

I like it..

I have all the same picks

good luck with your team Arkansas, but the only pick I don't like from the algorithm is the Arkansas pick. We will see what happens.

[cyptomcat]

so far pretty good

[cyptomcat]

6-0 for this week.

[cyptomcat]

Picks for this week:
Texas -1.5
Arkansas -14
Michigan pk
Western Kentucky +41.5
Nebraska -3
Kansas State +4.5

[cyptomcat]

4-1 this week with 1 push, pretty good!

[cyptomcat]

Didn't get good numbers from algorithm this week. 4 picks:
Ohio State -7
UVA +17.5
Stanford -17.5
Kansas State +8

[cyptomcat]

2-2 this week, unreliable with less than 7 picks

[cyptomcat]

all right, not many good picks from algorithm this week either, still:
Iowa +9.5
Wyoming +33
Auburn +21
Ohio State +7.5

[cyptomcat]

went 2-2 last week

This week's picks:
Michigan State +9.5
Oklahoma State -3.5
New Mexico +48.5
LSU -13.5
West Virginia -2.5

[cyptomcat]

good start for this week

[cyptomcat]

well that's a sweep, 5-0 this week

hit 5-team and 7-team parlays on 5dimes thanks to it

[cyptomcat]

algorithm is 72% for the season, you guys have any idea how I can market this?

[ncsubowen]

Something of a small sample size, anything can be marketed though.

Seems like the best thing to do would be to keep putting money on it. That's why I never understood the tout angle - if you're picking winners then aren't you already cashing?

[cyptomcat]

Something of a small sample size, anything can be marketed though.

Seems like the best thing to do would be to keep putting money on it. That's why I never understood the tout angle - if you're picking winners then aren't you already cashing?

yes I have played&cashed all of the ones above.

As far as back testing, I got a p-value of .05 that the picks are not coming from a .55 correct rate (chi-square test). As far as I understand, that means with 95% chance, the picks I had are better than 55% correct.

[ncsubowen]

How far have you backtested?

[cyptomcat]

How far have you backtested?

only this season, I would need to do more work to test further back. I have at least done a statistical test on the picks that I have mentioned above, but I understand that seasons can be statistically independent and my result not worth shit for a different season. I will do back test for previous seasons in the off season.

[ncsubowen]

Well, good job taking the effort. I thought I had something earlier this year but it stuck me just off 50% the rest of the season. The issue is that there are a lot of smart people out there betting, and chances are if you've thought of it, so have those people and the books have adjusted accordingly.

That's always the issue with mathematical betting models. They can work, but they are not easy and certainly what has worked in the past will not necessarily work in the future.

[cyptomcat]

UL Lafayette +6

[cyptomcat]

Good start for the bowl season, I should have one more pick this week and then several for 12/30 and 12/31.

Regular Season picks are 19-5.
Bowl Season picks are 1-0.

[cyptomcat]

La Tech +9.5

[cyptomcat]

And we move to 2-0 in the Bowl Season picks. Regular Season picks were 19-5.

Next pick should be this weekend.

[cyptomcat]

Missouri -4.5

[cyptomcat]

Well that's not a bad start for the game at all. I am curious if anyone has been using my picks at all.

[cyptomcat]

3-0 in bowls, 19-5 regular season.

Next two picks:
Notre Dame +3
Washington +10

[cyptomcat]

that was ugly to make the record 3-2.

Vanderbilt -1.5

[cyptomcat]

and the bowl slide continues!

Ohio State +2
Penn State +7

[cyptomcat]

Oklahoma State -4