In the coupon collector process, the goal is to assemble a collection of $n$ distinct coupons, while we get a random coupon at each time.
I am looking at a generalization of this problem, where at each time we get a batch of $k$ random coupons (with repetitions) at once, for some $k\in\mathbb N^+$.
Assume that we have already collected $N\le n$ distinct coupons and let $y\in\mathbb N$, what is the probability that we get $y$ new and distinct coupons in the next batch (i.e., we will have $N+y$ distinct coupons after that batch)?
Does this has a close-form formula?
For example, if $n=10, k=3$ and we have so far collected $N=6$ coupons, the probability of collecting another (exactly) $y=1$ coupon is
$$ (4/10)\cdot(7/10)^2 + (6/10)\cdot(4/10)\cdot (7/10) + (6/10)^2\cdot (4/10) = 0.508.$$
Here, I looked at it as if the samples in the batch were ordered 1,2,3, and the summands corresponds to getting a new coupon at the first/second/third sample.
This approach doesn't seem to allow computation over a reasonable number of arguments.
Is there a better way to evaluate it?