1: \begin{abstract}

2:   We study the Gilbert-Shannon-Reeds model for riffle shuffles and ask

3:   'How many times must a deck of cards be shuffled for the deck to be

4:   in close to random order?'. In 1992, Bayer and Diaconis gave a

5:   solution which gives exact and asymptotic results for all decks of

6:   practical interest, e.g. a deck of 52 cards. But what if one only

7:   cares about the colors of the cards or disregards the suits focusing

8:   solely on the ranks? More generally, how does the rate of

9:   convergence of a Markov chain change if we are interested in only

10:   certain features? Our exploration of this problem takes us through

11:   random walks on groups and their cosets, discovering along the way

12:   exact formulas leading to interesting combinatorics, an 'amazing

13:   matrix', and new analytic methods which produce a completely general

14:   asymptotic solution that is remarkable accurate.

15: \end{abstract}

16: