3 . 3 Card Shuffling

Given a deck of n cards, how many times must we shuffle it to make it “random”? Of course, the answer depends upon the method of shuffling which is used and what we mean by “random.” We shall begin the study of this question by considering a standard model for the riffle shuffle. We begin with a deck of n cards, which we will assume are labelled in increasing order with the integers from 1 to n. A riffle shuffle consists of a cut of the deck into two stacks and an interleaving of the two stacks. For example, if n = 6, the initial ordering is (1, 2, 3, 4, 5, 6), and a cut might occur between cards 2 and 3. This gives rise to two stacks, namely (1, 2) and (3, 4, 5, 6). These are interleaved to form a new ordering of the deck. For example, these two stacks might form the ordering (1, 3, 4, 2, 5, 6). In order to discuss such shuffles, we need to assign a probability distribution to the set of all possible shuffles. There are several reasonable ways in which this can be done. We will give several different assignment strategies, and show that they are equivalent. (This does not mean that this assignment is the only reasonable one.) First, we assign the binomial probability b(n, 1/2, k) to the event that the cut occurs after the kth card. Next, we assume that all possible interleavings, given a cut, are equally likely. Thus, to complete the assignment of probabilities, we need to determine the number of possible interleavings of two stacks of cards, with k and n− k cards, respectively. We begin by writing the second stack in a line, with spaces in between each pair of consecutive cards, and with spaces at the beginning and end (so there are n − k + 1 spaces). We choose, with replacement, k of these spaces, and place the cards from the first stack in the chosen spaces. This can be done in ( n k )