Suppose a “fair” die is rolled 100 times. Let $X_i$ be the value obtained in the i-th roll. Calculate an approximation for:

$P( \displaystyle\prod_{i=1}^{100} X_i \leq a^{100})$, where $1 < a < 6$

I know that $X_i$~Unif{1,2,3,4,5,6}, and I think I should use some kind of normal distribution to approximate the probability, but I don't know what to do with the product. Can someone help me?


Hint: try taking the log of both sides of the inequality.

$$P(\prod_{i=1}^{100} X_i \le a^{100}) = P(\frac{1}{100} \sum_{i=1}^n \log X_i \le \log a)$$ and $\frac{1}{100} \sum_{i=1}^n \log X_i$ is approximately normal (you need to find the mean and variance).

Mean: $E \frac{1}{100} \sum_{i=1}^{100} \log X_i = E \log X_1 = \frac{1}{6} \log 720.$

Variance: $$\text{Var}(\frac{1}{100} \sum_{i=1}^{100} \log X_i) = \frac{1}{100} \text{Var}(\log X_1) = \frac{1}{600} \sum_{x=1}^6 (\log x - \frac{1}{6} \log 720)^2.$$

  • $\begingroup$ But $Z = \frac{X-E(X)}{Var(X)}$, how does $\frac{1}{100} \sum_{i=1}^n \log X_i$ is equal to $Z$? $\endgroup$ – Mystery Nov 23 '19 at 21:16
  • $\begingroup$ @Mystery I only said that its distribution is approximately normal (not standard normal). If you find the mean and variance of $\frac{1}{100} \sum_{i=1}^n \log X_i$, then you can standardize it by subtracting the mean and dividing by the standard deviation (not the variance) to get a random variable that is approximately standard normal. $\endgroup$ – angryavian Nov 24 '19 at 0:11

Your Answer

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

Not the answer you're looking for? Browse other questions tagged or ask your own question.