2
$\begingroup$

I assume this is stats 101 level of question.

If you have PC that is set to "randomly" pick a number between 1 and 100 each time a button is pushed, the chances of getting a 50 is 1 in 100, right?

So if you have two PCs doing the exact same thing, and you can press a button on either, but you can only pick one PC, what are the chances of getting a 50?

Also if you have one person at each PC, each pressing the respective button, what are each of their chances that one of them will get a 50?

$\endgroup$
  • 1
    $\begingroup$ "Randomly pick a number" is not precise enough to allow this to be answered. If the computer picks a number from a discrete uniform distribution, then the odds of picking a 50 are indeed, 1 in 100. But if it picks a number from any continuous distribution, the odds of picking a 50 exactly are infinitesimal. If it picks from some other discrete distribution, the odds depend on the distribution. I suspect you meant the first thing. As for your other question: Do you mean exactly one person picking a 50? Or at least one person picking a 50? $\endgroup$ – Peter Flom - Reinstate Monica Aug 27 '13 at 19:16
  • $\begingroup$ Thanks for the clarifying questions, and the time you've taken to expose my flawed premise. I do mean 1 in 100 integers. and I mean , 1 person, taking one attempt at pressing the button. $\endgroup$ – beauk Aug 27 '13 at 21:40
  • $\begingroup$ The odds that any particular person will pick a 50 is 1 in 100. $\endgroup$ – Peter Flom - Reinstate Monica Aug 27 '13 at 21:59
  • 1
    $\begingroup$ @AncientAnt All the questions seem identical. 1. 1/100. 2. Same because who cares if there is another PC? 3. Again, why would you care about anyone pressing button on some other PC? They are independent. $\endgroup$ – sashkello Aug 27 '13 at 23:28
1
$\begingroup$

If you have PC that is set to "randomly" pick a number between 1 and 100 each time a button is pushed, the chances of getting a 50 is 1 in 100, right?

Assuming by "randomly pick" you mean "with equal probability", then yes.

So if you have two PCs doing the exact same thing, and you can press a button on either, but you can only pick one PC, what are the chances of getting a 50?

Yes. Law of total probability ($\Pr(A)=\sum_i \Pr(A| B_i)\Pr(B_i)$). Let's call getting 50 a success ("S") and let $C_A$ and $C_B$ be the events that you picked computer A or B respectively.

$P(S)=P(S|C_A)\cdot P(C_A)+P(S|C_B)\cdot P(C_B)$

$ \quad= \frac{1}{50}\cdot P(C_A)+\frac{1}{50}\cdot (1-P(C_A))$

$ \quad= \frac{1}{50} (P(C_A)+ (1-P(C_A)))$

$ \quad= \frac{1}{50}$

Also if you have one person at each PC, each pressing the respective button, what are each of their chances that one of them will get a 50?

Depends if you mean "exactly one" or "at least one". If the PC's work independently, then

a) $P(\text{At least one }S) = 1-P(\text{no } S) = 1-(\frac{49}{50})^2 = 0.0396$

(i.e. just under 2/50)

b) $P(\text{Exactly one }S) = P(\text{success on }A, \text{fail on }B) + P( \text{fail on }A, \text{success on }B)$
$\quad = \frac{1}{50}\cdot \frac{49}{50} + \frac{49}{50}\cdot \frac{1}{50}$
$\quad = 0.0392$

Alternate method:

$P(\text{Exactly one }S) = P(\text{At least one }S)-P(\text{Both }S)$
$\quad = 0.0396 - \frac{1}{50}\cdot \frac{1}{50}$
$\quad = 0.0392$

$\endgroup$

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.