I expect lots of brown M&M's - my theory of the candy is that you need a large base of browns to establish the essential nature of the chocolate experience, but the colors add a variety that can be exploited either by an extensive planning process that consumes the colors in a balanced sequence, or in which one competes with siblings for the rare colors, simultaneously belittling the intelligence of anyone who gets ahead (if someone grabs two blues, you say that blues are yucky and grab two greens, declaring them to be superior). If you select m&m's at random from the bag, how close do you expect to get to a balanced sequence? What is a balanced sequence? Intuitively, it means that any contiguous subsequence has the expected number of colors. If there are 4 blues out of 60 candies, you'd want them spaced out about 15 apart. So, I have never believed that the colors were supposed to drawn from a uniform distribution. There should be more browns. Beyond that, I've suspected that the rest were indeed drawn from a uniform distribution. I think your numbers support that, except that blues are indeed new, and are meant to incite envy. I don't know if blues were taken out of the green distribution or added as a new category. As I remember one of my favorite urn problems, if you are trying to distinguish between two possible distributions, sampling is a very accurate method. But, it you don't know anything about the distribution, then you need a lot of samples. At 240 calories per bag, maybe you don't want to know the answer. Hilarie