How frequently do Skittles packs contain an identical payload?

Originally published at:


Using the term “payload” to describe an amount of food, even one as processed as Skittles, makes me feel strangely uneasy…


Reading this story makes my teeth hurt…


I . . . I feel validated.

Alas, since retiring, I know longer have the luxury of getting a daily packet of M&Ms from the office candy jar, which was a great way to get lots of samples, and I’m trying to lose weight so my quest to entirely tally a 5 lb. bag is off the table.

I guess I could sort and tally a bag and then put the candies in a nice container and leave them at place where I volunteer.


When I was in college, I used a 2 pound bag of m&ms as a study reward for my physics final.

after dumping out the bag, was astonished to discover that there were no green ones!

I wrote a strongly worded note to M&Ms, and was amazed to discover that I got coupons for 6 pounds of free replacements in the mail.

go figure. I was not alone in my color counting.


I don’t think I would want to calculate the exact PDF. :cold_sweat:

1 Like

Oscar Charlie Delta.


He could have built a LEGO Technic sorter, increased sample size to a million skittles, and still finished in less than 3 months.



My oldest daughter got a Skittle stuck in her nose when she was two. It’s a damned good thing we didn’t have 27,739 other Skittles lying around, or there’s no telling how full her head would have gotten.


there’s a fair chance you’ll get anywhere between 55 and 65 (though in fairness there’s about a 2 in 3 chance of getting 59,60 or 61)

Presumably they only care about weight - and also presumably carefully control that, to avoid giving too much candy*, which suggests the individual candies vary in size/density quite a bit.

*Although, alternately, this might just mean that 65-ish candies is the target and they consistently shortchange buyers.

1 Like

+1 for fun.

-1 Because statistically, this is bullshit with respect to @beschizza 's title: it does not, for example, account for replicating the sample distribution of Skittles™ in the second bag, or the sample distribution of the third bag, etc.


I believe you mean…
A = 3, C = 2, D = 1, E = 2, H = 1, I = 1, L = 2, O = 1, R = 2, S = 1, T = 1



It’s not all THAT difficult. Start by assuming, for the heck of it, that all colors are equiprobable, and posit that, to avoid angry consumers, they make sure at least N of each color get into every packet. You can ignore the variation in package count by “binning” each packet count value separately.

Friday’s a holiday, so have the answer ready to turn in at the start of Monday’s class.

Noooooo! :tired_face: Nine characters of noooooooo!

1 Like

Okay. Now I’d like to know how the individual Skittles differ in weight, and if there is a significant spread across the different colours.

The gold-coated ones are obviously heaviest.

What? You don’t have gold-coated Skittles?

I can’t help you.


I can only imagine that the green hopper had a jam at the time that 2 lb. bag was filled.

Come to think about it, I’m not sure it would work out that way; gold is so malleable that coatings/plating could be incredible thin…
Further testing is needed.