# Calculating probability of $(n_1,n_2,n_3,\dots, n_k)$-of-a-hands for a generic deck of cards.

Let's say I have a generalized deck of cards, consisting of $R$ ranks, $S$ suits, and with $C$ copies of each card. The number of cards in the full deck would thus be $RSC$, and the number of cards of each rank would be $SC$. In addition, I have a hand size of $H$.

I'm trying to find a general equation for the probability of getting dealt a generic $(N_1, N_2, ..., N_k)$-of-a-kind in a hand of size $H$. An example of such a hand would be a classic two pair hand in standard poker (e.g. 4, 4, K, K, ?). This would be a $(2,2)$-of-a-kind, with $H=5$.

I've tried going about this problem by attempting to calculate the number of combinations for a fixed set of givens. The following equation for number of combinations resulted from a quick attempt at the problem, but it doesn't seem to be correct:

$$\left({\prod^k_{i=1}\left(R-i+1\right)\binom{SC}{N_i}}\right)\binom{R-k}{H-W}{SC}^{H-W}$$

Given $(N_1, N_2, ..., N_k)$ and $W=\sum^k_{i=1}N_i$.

I must be overlooking something; probability and stats are not my strong points, for sure. Any assistance as to what I should be doing?

-
Why doesn't your formula seem to be correct? Can you illustrate with a simple example? – oks Mar 1 '13 at 10:27
I don't see why you're differentiating between suits and copies. That distinction doesn't enter into the probability you're looking for; your answer depends only on $SC$, as the correct answer must. – joriki Mar 1 '13 at 10:48

Say you want $k_j$ $j$-tuples of cards of the same rank, i.e. $k_1$ singletons, $k_2$ pairs, etc., with $\sum_jjk_j=H$; then the probability is
$$\binom{RSC}H^{-1}\pmatrix{R\\k_1,\dotsc,k_n,R-\sum_jk_j}\prod_j\binom{SC}j^{k_j}\;,$$
$$\pmatrix{R\\k_1,\dotsc,k_n,R-\sum_jk_j}=\frac{R!}{k_1!\cdots k_n!\left(R-\sum_jk_j\right)!}$$
that counts the number of ways of choosing $k_1$ ranks for the singletons, $k_2$ ranks for the pairs, etc.