[SlickRick1382]

Can someone link me to an accurate article, website, etc. that can properly show me how to calculate the probability or percentage of an event occurring based on various different percentages. Or even one where the percentages can be weighted differently.

Simple example not relating to what I'm trying to do but shows 4 different variables that would need to be computed.

What's the probability that Team A or B scores a run in the 1st inning if Team A normally scores a run 25 % of the time and allows its opposition to score 30% of the time and Team B scores 27% of the time and allows opponents to score 23% of the time.

Thanks in advance

[James Marques]

http://en.wikipedia.org/wiki/Bayes'_theorem

[SlickRick1382]

James Marques

Thank you for the link and the help.

If I were to post the above sample and solve it using Bayes' theorem would you be able to tell me if I'm structuring it correctly? Just to make sure I properly understand it before proceeding... That's assuming you have an understanding of it

Thanks in advance ....

[James Marques]

Yes, I can help. I'm pretty solid on Bayes and conditional probability as a whole.

[buby74]

Don't you also need the league average as the answer will be different if the average is 20% rather than 35%

[SlickRick1382]

Correct, league averages, home and away averages if we want to be specific, etc.

It was just a quick simple example in order to point me in the right direction.

[James Marques]

No... careful guys.... League MEDIANS! Very important distinction.

[James Marques]

Slick got your PM. We'll discuss.

[KVB]

No... careful guys.... League MEDIANS! Very important distinction.

Hey James, I owe you a message but just haven't formulated the steps to show you, been busy, excuse, excuse. I'll be in touch.

Anyway, James is right, stick with medians, averages will screw you up in most, not all, but most cases when comparing teams to the league.

[SlickRick1382]

No... careful guys.... League MEDIANS! Very important distinction.

Correct, I meant to post median not mean but didn't notice the mistake. One large enough variation in the average can distort the numbers.

Thanks for the reply, really appreciate it bud.

Slick got your PM. We'll discuss.

Sounds good. I'll look out for it and in the interim work on what info to include so you can let me know if I'm correct in my assessment

[statnerds]

keep going and add another 12 variables and indicators and then you should begin to arrive at the numbers formulated by better modelers, market shapers, and whatnots.

that isn't too say that what you are doing is worthless as long term it can only improve your modeling abilities.

having said that, unless this Market has large max bets, and assuming you will be successful at it, you will quickly reach a point where it isn't worth your time.

so for the future practical applications, go for it. as for thinking it will be profitable long term I only caution that it may turn out that way, and that every angle/indicator you can think of is already accounted for in an efficient market.

[lamichaeljames]

keep going and add another 12 variables and indicators and then you should begin to arrive at the numbers formulated by better modelers, market shapers, and whatnots.

that isn't too say that what you are doing is worthless as long term it can only improve your modeling abilities.

having said that, unless this Market has large max bets, and assuming you will be successful at it, you will quickly reach a point where it isn't worth your time.

so for the future practical applications, go for it. as for thinking it will be profitable long term I only caution that it may turn out that way, and that every angle/indicator you can think of is already accounted for in an efficient market.

I would argue that the market is not as efficient as one might think. There are advantages in betting lines early.

[KVB]

I would argue that the market is not as efficient as one might think. There are advantages in betting lines early.

Indeed.

[magyarsvensk]

Probabilities are a bitch to deal with all around. Move away from probabilities and towards historical simulation. It's simpler to calculate and more versatile.