how many people here use models to bet?

[pico]

you think your model is better than linemakers?

[SlickFazzer]

Linesmakers models are razor sharp in this day and age.

[RageWizard]

I don't know any models but if I did, I would surely let them bet for me because lady luck is nice to have on your side.

[jackpot269]

If I had models I be doing something besides betting

[Mudcat]

Not positive I understand. If by models you mean systems, I do make money with them.

Asking if my systems are better than the linesmakers is comparing apples and oranges. We are trying to accomplish different things. The linesmaker's goal is to get balanced action. My goal is to pick out angles where the action, which is a function of public psychology, has forced linesmakers to move a line out of whack from the true expectation of the proposition.

In many ways, the linesmaker and I are working together.

Of course linesmakers can have totally different styles and skills so in a way this question is like asking. "Are you smarter than someone at the mall?"

Depends who.

[Justin7]

Why would you bet without knowing what the right side is? Of course everyone uses models!

[durito]

Linesmakers models are razor sharp in this day and age.

then why do the openers move so much?

[pico]

for those you who use models, do you use weighted average or regression forecast models? or something else?

[Justin7]

for those you who use models, do you use weighted average or regression forecast models? or something else?

I use regression to weigh different factors... But the models are always result-driven - e.g. predicting a final score, and distribution around that score.

[pico]

I use regression to weigh different factors... But the models are always result-driven - e.g. predicting a final score, and distribution around that score.

justin, do you just use simple linear regression models with montel carlo simulation? how many variables do you typically use in a ncaa fb game?

[Mudcat]

for those you who use models, do you use weighted average or regression forecast models? or something else?

I don't know. I don't use that terminology. I'm not criticizing it; I just don't know quite what it means.

If I had to apply a terminology to my "model", I might call it the BLUNT FORCE method. I research the hell out of bets offered by Matchbook (because their lines are so insanely good), gathering and crunching numbers, then I find angles where the expected outcome is out of whack from the odds based on a very large number of samples and then I beat the hell out of it with a stick.

I'm being a bit silly of course. There is a fair bit of finesse and expertise involved in deciding what angles to investigate and in the line shopping (because, although Matchbook is the cornerstone, knowing what a lot of books offer is critical).

[Justin7]

justin, do you just use simple linear regression models with montel carlo simulation? how many variables do you typically use in a ncaa fb game?

No monte carlo. For NCAA, I track about 20 variables per game. I only use games of the current year to evaluate a team.

[pats3peat]

i hope to implement some

[LLXC]

Target would be to make a model that is very close to the linemakers and take whatever points you can get (cause you know they shave for the public).

[donjuan]

Target would be to make a model that is very close to the linemakers and take whatever points you can get (cause you know they shave for the public).

Fail.