[TPowell]

If I calculated the conference wins for teams in the ACC, how would I be able to price the teams to the futures bet? (which team wins the conference).

[TPowell]

anybody? I feel like I should know how to do this but I'm blank

[tto827]

If you are looking at making a futures bet to win the ACC you have still got a lot of math ahead of you. You then have to use those numbers to figure out the probability that each team ends up with the most wins in the conference. The shortest/easiest/least effective way, would be to calculate how many wins you think it will take to win the ACC this year, then determine the probability that each team gets to that number of wins. That can be done with basic stats knowledge. Someone who deals with stats more regularly then I do is going to have to help you with how you would calculate the chances of each team beating all other teams based on their projected wins.

[TPowell]

If you are looking at making a futures bet to win the ACC you have still got a lot of math ahead of you. You then have to use those numbers to figure out the probability that each team ends up with the most wins in the conference. The shortest/easiest/least effective way, would be to calculate how many wins you think it will take to win the ACC this year, then determine the probability that each team gets to that number of wins. That can be done with basic stats knowledge. Someone who deals with stats more regularly then I do is going to have to help you with how you would calculate the chances of each team beating all other teams based on their projected wins.

I have projected conference wins and I can obviously use an arbitrary number as the wins needed to win the conference. I fail to understand how to determine the probability of teams getting to that number though. Wish they had season wins out for the teams instead

[TPowell]

been doing things with pen and paper, but have access to Excel which I'm sure I'll need to calculate the probability of teams getting to the number of wins it takes to win the conference. Just not sure what exactly i would use. The kids in my entry level stats class were clueless so we didn't get very far

[tto827]

I have projected conference wins and I can obviously use an arbitrary number as the wins needed to win the conference. I fail to understand how to determine the probability of teams getting to that number though. Wish they had season wins out for the teams instead

Gimme a couple minutes, maybe half an hour to touch back up on my statistics and I'll get back to you with how to get to that number.

[TPowell]

Gimme a couple minutes, maybe half an hour to touch back up on my statistics and I'll get back to you with how to get to that number.

appreciate it pal. I'll be on my phone more than likely checking SBR then but I can test it out tomorrow

[tto827]

I'm getting pretty close, it'll make it a lot easier to explain over the internet if you give me the arbitrary number before I work out the math. (I can also show you how to work it out with a different arbitrary number, but lets make life easy to start)

[TPowell]

I'm getting pretty close, it'll make it a lot easier to explain over the internet if you give me the arbitrary number before I work out the math. (I can also show you how to work it out with a different arbitrary number, but lets make life easy to start)

13 is my number

[tto827]

13 is my number

cool, as much as I wanted to solve this tonight, (its gonna be haunting me in my sleep). I will try and have it solved by like 12pm Pacific time tomorrow (today I guess)

[TPowell]

cool, as much as I wanted to solve this tonight, (its gonna be haunting me in my sleep). I will try and have it solved by like 12pm Pacific time tomorrow (today I guess)

No rush man, appreciate the help

[hutennis]

I guess, what tto827 is trying to say is that problem you face is enormously difficult and it takes huge resources (time, knowledge, money)
to pin it down. I remember Bob Voulgaris talking about spending upwards of 3 mil dollars over a few years on his R&D.
This kinda explains why only tiny fraction of a participants can beat this game long term. It is not easy.
If it would be possible to find help you need for free just by asking for a little help in a public internet forum this industry would not exists.

Moreover, the numbers you are looking for are already there and readily available.
You don't need to do anything. Somebody has already done all your work for you.
All you need is to convert futures odds into implied probabilities and you have it.
If you think that numbers you'll get are close enough, then act accordingly.
If you think that those who've created these numbers are less competent than you are and you want to argue with them - then it is fair enough too. That is what Bob Voulgaris does. Successfully, I might add.
But if it takes him upwards of 3 mil dollars and a few years, what makes you think you can do it for close to nothing by next week?

[tto827]

I guess, what tto827 is trying to say is that problem you face is enormously difficult and it takes huge resources (time, knowledge, money)
to pin it down. I remember Bob Voulgaris talking about spending upwards of 3 mil dollars over a few years on his R&D.
This kinda explains why only tiny fraction of a participants can beat this game long term. It is not easy.
If it would be possible to find help you need for free just by asking for a little help in a public internet forum this industry would not exists.

Moreover, the numbers you are looking for are already there and readily available.
You don't need to do anything. Somebody has already done all your work for you.
All you need is to convert futures odds into implied probabilities and you have it.
If you think that numbers you'll get are close enough, then act accordingly.
If you think that those who've created these numbers are less competent than you are and you want to argue with them - then it is fair enough too. That is what Bob Voulgaris does. Successfully, I might add.
But if it takes him upwards of 3 mil dollars and a few years, what makes you think you can do it for close to nothing by next week?

Great post, anyone with some stats background want to double check that this would work, cause its the only way I've come up with that will give the likelihood of of winning the conference (in any sort of reasonable amount of time).

You said 13 games to win the conference, so we have to add P(13)+P(14)+P(15)+P(16)+P(17)+P(18) for each team. (I am pretty sure I am still accurate here)

Now for this example lets say I project Duke to win 12 games, that is a .667 win rate.

And finally the magical equation to solve all your problems... Will be included when I turn my computer on because my iPhone wont do sub/super scripts.

[tto827]

C_xⁿ is a combination

n = the number of trials (18 in this case, # of conference games each team plays)

x = number of wins (must use 13, 14, 15, 16, 17, and 18, then add up results)

p = probability of success in a single trial (win percentage of each team for any given game, total expected wins/ total games)

q = the probability of failure in a single trial (1-p)


edit: we're good now

P(x)= C_xⁿ * p^x * q^n-x

[TPowell]

Thanks pal, for some reason when something doesn't click when I try to do it, I just can't pull it from my head. Makes sense

[TPowell]

I guess, what tto827 is trying to say is that problem you face is enormously difficult and it takes huge resources (time, knowledge, money)
to pin it down. I remember Bob Voulgaris talking about spending upwards of 3 mil dollars over a few years on his R&D.
This kinda explains why only tiny fraction of a participants can beat this game long term. It is not easy.
If it would be possible to find help you need for free just by asking for a little help in a public internet forum this industry would not exists.

Moreover, the numbers you are looking for are already there and readily available.
You don't need to do anything. Somebody has already done all your work for you.
All you need is to convert futures odds into implied probabilities and you have it.
If you think that numbers you'll get are close enough, then act accordingly.
If you think that those who've created these numbers are less competent than you are and you want to argue with them - then it is fair enough too. That is what Bob Voulgaris does. Successfully, I might add.
But if it takes him upwards of 3 mil dollars and a few years, what makes you think you can do it for close to nothing by next week?

If I took 3 hours preparing a sandwich, do you think it would be better than yours if you took 10 minutes? The main thing is the underlying assumptions imo

[tto827]

Thanks pal, for some reason when something doesn't click when I try to do it, I just can't pull it from my head. Makes sense

I sat with a pen and paper for hours dude, it was driving me insane. Let me know if you find anything with significant value please.

[hutennis]

If I took 3 hours preparing a sandwich, do you think it would be better than yours if you took 10 minutes? The main thing is the underlying assumptions imo

In your opinion, does it take about the same set of skills and comparable resources to make a sandwich as to correctly estimate probabilities of multiple future events with countless variables to consider?

[TPowell]

In your opinion, does it take about the same set of skills and comparable resources to make a sandwich as to correctly estimate probabilities of multiple future events with countless variables to consider?

touche

[hutennis]

touche

Good for you.

[tto827]

Gonna bump this cause the equation that I attempted to use did not work. Can anyone give us an equation or tell me where I went wrong in post 14?