I have always been struggling with the problem, in particular, I usually have great difficulty in differenting when should I multiply n! to take care of the ordering, and when should I not do so. For simple question, this might not be an issue; but if this basic concept is not understood throughly, this will extend to more complicate question. Because of this, I hope you guys can correct my understanding of basic probably.

For instance, a very basic question:

12 balls in a pool, 7 green, 2 yellow, 3 orange. What is the probability that of drawing 2 yellow and 2 orange?

My attempt:

First Approach:

the simplest way is to use view each draw is independent (this approach I have no problem): $$ \frac{3}{12}\cdot\frac{2}{11}\cdot\frac{2}{10}\cdot\frac{1}{9}\cdot4! $$ 4! because such "independent draw" could happen 4! times. So this approach I have no problem.

Second Approach: \begin{align*} P(\text{2 yellow AND 2 orange})&=\frac{\text{# of ways you can draw 2 yellow AND 2 orange}}{\text{# of ways draw 4 balls}} \\ &=\frac{\binom{3}{2}\binom{2}{2}}{\binom{12}{4}} \end{align*}

my understanding:

  • $\binom{3}{2}$ is the number of ways you can form 2 orange out of the 3, where the ordering is not important.
  • you mutiply by $\binom{2}{2}$ is because for each of $\binom{3}{2}$ ways, you can have $\binom{2}{2}$ ways to combine them with the yellow balls.

It terms out that I am missing a factor of 4. It turns out that I am having true difficulty with this type of problem. What can I do to improve?


If you count the number of ways of drawing the balls such that balls of the same color are distinguishable and such that order does not matter, then you should have $$\frac{\binom{3}{2} \cdot 1}{\binom{12}{4}},$$ as you have written.

The first approach is incorrect. $\frac{3}{12} \cdot \frac{2}{11} \cdot \frac{2}{10} \cdot \frac{1}{9}$ is the probability of choosing orange first, orange second, yellow third, and yellow fourth, as well as any other ordering. You only want to multiply this by $\frac{4}{2!2!}$ (the number of ways of arranging "orange, orange, yellow, yellow" in a row) instead of $4!$ (number of permutations of 4 distinguishable objects). The erroneous factor of $4$ comes because each arrangement of "orange, orange, yellow, yellow" is counted 4 times instead of once because you distinguished between the two oranges and the two yellows when multiplying by $4!$.

  • $\begingroup$ yes, thank you very much, can I ask you further question? First of all, I agree with your answer. In short, you are saying that OYOY and OOYY is distinct and you dont differentiate O1 and O2 or Y1 and Y2 and Y3, etc $\endgroup$ – kuku Sep 14 '14 at 4:35
  • $\begingroup$ Second, I am having question on (3C2) term. (3C2) means the number of ways you can form a group of 2 from 3 distinct objects. So that, you are treating each Orange balls (O1, O2, O3) as distinct objects, and you are seaking the number of groups of 2 you can form from 3 distincts Orange balls. (eg, O1O2, O1O3, O2O3). So here is my confusion: in my first comment, where you don't differentiate objects of same color, but now, you are differentiate them in (3C2) term. And I know this must be a misunderstanding from me. Can you please explain? I think the answer to this question is my key concern $\endgroup$ – kuku Sep 14 '14 at 4:41
  • $\begingroup$ @kuku As I wrote above, the approach with the $\binom{3}{2}$ involves the balls of the same color being distinguishable. This is ok as long as you assume the same thing when computing the denominator. $\endgroup$ – angryavian Sep 14 '14 at 11:15

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.