Suppose there are $500$ people in our office. I want to find the probability of drawing my friend's name. Rules: every person picks out a name from a hat without replacement and for this problem, it's fine to draw one's own name if that simplifies calculation. (Alternatively, calculation may be simpler if we disallow picking one's own name - that's fair game too - whatever makes our life easier)
My approach is to model the probability of being the $k$-th person to draw, or $Q_k$, and the probability of the $k$-th person drawing my friend's name, or $P_k$, via a recurrence relation: $$P_k = (1-P_1)(1-P_2)\ldots (1-P_{k-1})\bigg(\frac{1}{500-k}\bigg)$$
$$Q_k=(1-Q_1)(1-Q_2)\ldots(1-Q_{k-1})\bigg(\frac{1}{500-k+1}\bigg)$$
Then the answer can be expressed as $\sum P_k Q_k$.
But this doesn't seem to result in a nice answer. Is there something I'm missing in this method? Is there an alternative method instead?