Am working on a probability problem where a game is described using the hypergeometric distribution. The sample successes are $3...7$, the number of samples is $7$, population successes are $10$, and the number in population is $55$. Summing the probabilities gives $\sim0.1040683327$ or "overall odds" of about $1$ in $9.61$.

Here's the tricky part. For each play, there are $3$ sets of $7$ numbers, and only one set may be chosen by the player (the other $2$ are chosen randomly). Basically, $3$ independent trials on a single slip. So how do you calculate the overall odds of winning per play for each level ($3...7$)? According to the operator it is $1$ in $3.6$. Using the Binomial distribution, taking the probability of at least one success and averaging it with the probability of exactly one success gives overall odds of exactly $1$ in $3.6$. However, this is not the closest approximation compared to using the Rule of Complements on the sum of probability from the hypergeometric distribution mentioned earlier. That result comes out to be $\sim 0.2808414272$ or about $1$ in $3.56$. The closest approximation seems to come from using a formula derived from the Poisson distribution...this gives overall odds of $1$ in $3.58$. Keep in mind I am trying to calculate the closest probability of each success level so that the overall odds come out to be nearest $1$ in $3.6$. Now, rounded to a couple decimal places both methods I tried come to the same results, but I just want to make sure I am doing this correctly.

If this isn't clear, basically I want to calculate the actual probability of winning for each tier based on the fact that 3 trials occur in the same play. Obviously it isn't as simple as dividing each odds by 3 since there is the chance of winning more than once.



Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.