Question

Why do we have to use "combinations of n things taken x at a time" when we calculate binomial probabilities?

Answer 1

See below on my thoughts:

Explanation:

The general form for a binomial probability is:

`sum_(k=0)^(n)C_(n,k)(p)^k((~p)^(n-k))`

The question is Why do we need that first term, the combination term?

Let's work an example and then it'll come clear.

Let's look at the binomial probability of flipping a coin 3 times. Let's set getting heads to be `p` and of not getting heads `~p` (both `=1/2)`.

When we go through the summation process, the 4 terms of the summation will equal 1 (in essence, we are finding all the possible outcomes and so the probability of all the outcomes summed up is 1):

`sum_(k=0)^(3)=color(red)(C_(3,0)(1/2)^0((1/2)^(3)))+color(blue)(C_(3,1)(1/2)^1((1/2)^(2)))+C_(3,2)(1/2)^2((1/2)^(1))+C_(3,3)(1/2)^3((1/2)^(0))`

So let's talk about the red term and the blue term.

The red term describes the results of getting 3 tails. There is only 1 way for that to be achieved, and so we have a combination that equals 1.

Note that the last term, the one describing getting all heads, also has a combination that equals 1 because again there is only one way to achieve it.

The blue term describes the results of getting 2 tails and 1 head. There are 3 ways that can happen: TTH, THT, HTT. And so we have a combination that equals 3.

Note that the third term describes getting 1 tails and 2 heads and again there are 3 ways to achieve that and so the combination equals 3.

In fact, in any binomial distribution, we have to find the probability of a single kind of event, such as the probability of achieving 2 heads and 1 tails, and then multiplying it by the number of ways it can be achieved. Since we don't care about the order in which the results are achieved, we use a combination formula (and not, say, a permutation formula).