Let me explain what I mean. M&Ms are just the kind of treat that invite adventures in numerology. There are a variable number of candies in each bag, and the distribution of colors in each bag is variable, although statistically predictable over the long term. For someone interested in numbers, it is natural to separate the contents of a bag into piles by color and count 'em up. It is perhaps a bit unnatural to track those counts on paper, but that's what I do.
In early 1998, I purchased a case of M&M bags from the local Costco. Having an entire case made for an excellent sample population, so I started tracking the color counts, and even produced the occasional Excel pie chart to show the sample distributions. Harmless fun. Several of my partners and co-workers chuckled at my eccentricity, but that's the price we pay.
About halfway through the case, I started noticing some statistical anomalies. Ordinarily, the average mid-sized bag of M&Ms contains about 54 candies, plus or minus 2, with the 6 colors in a relatively predictable distribution. One day, I opened a bag and counted 63 candies, in a startling 1-2-4-8-16-32 distribution. This was a six-sigma event, and I was quite tickled for having seen it. The next day, I opened a bag and again counted 63 candies, this time in a 32-16-8-4-2-1 distribution. Naturally, I shared this once-in-a-lifetime alignment with my office mates, who had a good laugh at my expense. (Again.)
The following week, I encountered a bag with no browns at all. Since dark brown is by far the most common color, this was astounding. Later that same week, I opened a bag with a 18-18-18-0-0-0 distribution, and a matching 0-0-0-18-18-18 bag the week after. I began to wonder if someone at the factory was playing games with me.
A week after that, I opened a bag that was half regular M&M's, and half mini-M&Ms. Again, I marvelled at the anomaly, and wondered what had gone wrong with quality control at Mars, Inc.
The remainder of the case was uneventful, and in time I forgot about it. That is, until the annual company Christmas party in December. Here, my co-workers unveiled a horrible secret that changed my life forever. At a well-lubricated party in mid-year, a co-worker discovered that it was possible to open a bag of M&Ms and then reseal it. The resealed bag would never pass a professional inspection, but to the casual snacker, it looked quite normal. They had absconded with a number of my bags of candy, opened them, dumped the contents onto a table along with a jumbo bag for seed, and refilled my bags with those anomalous counts, knowing that I was taking inventory. The rest of the office had been told that, if they took a bag of my M&Ms, to be sure to grab from the bottom (untampered) part of the case.
I was stunned. I had no clue that I was being scammed. Looking back, I can remember sharing my disbelief with a co-worker, who stifled a laugh behind his hands. I assumed he was laughing at my enthusiasm for numbers, rather than laughing at my gullibility.
And, of course, this makes a great cocktail party story. At business events, I have been introduced to people I have never met, and been greeted with "Are you the guy with the M&Ms?"
The lesson from all of this is "never trust a twin."
