Empirical Observations on M&Ms™

This page is unaffiliated with Mars, Inc.'s official "M&Ms"™ home page, www.m-ms.com.

Fact #1: Plain M&Ms™ are my candy of choice (aside from Junior Mints™, which by a happy quirk of fate are made right in my neighborhood. But I digress.).

Fact #2: Call me crazy, call me borderline-OCD, but I like to eat my M&Ms™ by color.*

Fact #3: Although the reported distribution has since changed, in 2000-2001 Mars, Inc. claimed that the colors in a bag of plain (excuse me, "milk chocolate") M&Ms™ were distributed as follows:

This was clearly specious. I decided to conduct a rigorous empirical observation to satisfy myself as to the real statistics.

Methodology: Each sample consisted of one (1) standard 1.69 oz/47.9g bag of plain M&Ms™. Samples were taken at random over a period of 365 days. Recorded percentages have been rounded to the nearest tenth of a percent.

How many of each color?
Date Brown Yellow Green Orange Red Blue Total
7/24/2000 22 37.9% 17 29.3% 6 10.3% 5 8.6% 6 10.3% 2 3.4% 58
7/25/2000 14 25.5% 14 25.5% 6 10.9% 9 16.4% 9 16.4% 3 5.2% 55
7/26/2000 10 18.2% 17 30.9% 8 14.5% 2 3.6% 10 18.2% 8 14.5% 55
7/27/2000 15 26.3% 11 19.3% 9 15.8% 7 12.3% 9 15.8% 6 10.5% 57
7/28/2000 20 36.4% 10 18.2% 6 10.9% 5 9.1% 8 14.5% 6 10.9% 55
8/7/2000 21 38.2% 7 12.7% 7 12.7% 3 5.5% 13 23.6% 4 7.3% 55
8/14/2000 20 37.0% 10 18.2% 3 5.6% 6 11.1% 10 18.5% 5 9.3% 54
8/22/2000 15 27.3% 10 18.2% 6 10.9% 7 12.7% 15 27.3% 2 3.6% 55
8/23/2000 21 38.2% 9 16.4% 0 0% 9 16.4% 13 23.6% 3 5.5% 55
8/24/2000 26 46.4% 12 21.4% 3 5.4% 2 3.6% 13 23.2% 0 0% 56
10/5/2000 1 11 19.6% 5 8.9% 10 17.9% 7 12.5% 11 19.6% 12 21.4% 56
10/5/2000 16 29.6% 7 13.0% 11 20.4% 5 9.3% 14 25.9% 1 1.9% 54
11/8/2000 17 30.4% 11 19.6% 5 8.9% 4 7.1% 13 23.2% 6 10.7% 56
11/9/2000 25 4.6% 5 8.9% 4 7.1% 6 10.7% 14 25.0% 2 3.6% 56
11/20/2000 2 15 27.8% 12 22.2% 4 7.4% 5 9.3% 11 20.4% 7 13.0% 54
1/29/2001 13 23.6% 2 3.6% 11 20.0% 6 10.9% 10 18.2% 13 23.6% 55
1/31/2001 19 35.8% 11 20.8% 5 9.4% 6 11.3% 9 17.0% 3 5.7% 53
2/9/2001 14 25.9% 8 14.8% 5 9.3% 11 20.4% 11 20.4% 5 9.3% 54
2/14/2001 18 32.7% 9 16.4% 4 7.3% 6 10.9% 11 20.0% 7 12.7% 55
2/21/2001 18 32.1% 11 19.6% 8 14.3% 2 3.6% 11 19.6% 6 10.7% 56
2/23/2001 20 35.7% 10 17.9% 6 10.7% 4 7.1% 10 17.9% 6 10.7% 56
2/26/2001 19 34.5% 7 12.7% 7 12.7% 6 10.9% 13 23.6% 3 5.5% 55
3/21/2001 20 35.7% 11 19.6% 3 5.4% 5 8.9% 15 26.8% 2 3.6% 56
4/6/2001 24 42.1% 16 28.1% 3 5.3% 5 8.8% 6 10.5% 3 5.3% 57
5/23/2001 20 35.1% 10 17.5% 6 10.5% 3 5.3% 15 26.3% 3 5.3% 57
6/11/2001 21 37.5% 5 8.9% 11 19.6% 5 8.9% 11 19.6% 3 5.4% 56
7/16/2001 14 25.9% 5 9.3% 9 16.7% 11 20.4% 12 22.2% 3 5.6% 54
7/24/2001 12 21.1% 9 15.8% 9 15.8% 13 22.8% 10 17.5% 4 7.0% 57
Cumulative Total &
Average Percentage
500 32.2% 271 17.5% 175 11.3% 165 10.6% 313 20.2% 128 8.2% 1,552
Theoretical Percentage 30 20 10 10 20 10
1 Contributed by Matthew Morse.
2 Contributed by Eric Mumpower.

Conclusions: Well, what do you know. They're not as far wrong as I originally thought from examining one bag at a time. Which goes to show the power of the sampling error. When I learn something about statistics, maybe I'll get into the mathematical analysis of this set of data, but for now, let's leave that as an exercise for the reader.


Q: What do you get when you combine one bag of M&Ms™ with one bag of Skittles™?
A: S&M's! HAHAHA!


* Circumstances permitting, of course. Generally, this means when I'm sitting at my computer at work and can spread them out on my desk. It amuses me; I'm not compulsive about it.
Send E-mail | Back to Erica's Home Page