The random check, and how rarely it is really random

Something was inspected. There was no reason available, nothing about it stood out, and it simply came up in a sample.

This does happen. Systems that handle huge volumes pull items at random because random sampling is how you catch what targeting misses. Being in that slice is not caused by anybody.

Random and selected are different words

A random check picks without reference to anything about the item. A selected check picks because something about the item invited attention. Both are described afterwards as random, by everybody involved, because nobody wants the second explanation.

The difference is not detectable from the outcome. It is detectable from the item, and the item is the part people do not want to examine.

Things that turn random into selected

• Weight or shape that does not match what the outside implies.
• A route or origin that gets more attention than average.
• Packaging that is careful in an unusual way, which is its own kind of signal.
• A sender or a service level that does not fit the rest of the picture.
• Anything detectable from outside the packaging, which is a preparation matter and belongs with the packing page.

None of these guarantees attention. All of them shift the odds, and shifted odds are not randomness.

Why the random story is so popular

Because it ends the conversation without anybody having to change anything. Random is unanswerable. It cannot be argued with, it cannot be improved on, and it lets the same method be used again next week.

That is what makes it worth interrogating. An explanation that recommends no change and permits everything to continue is the explanation most likely to be reached for whether or not it is true.

What follows from getting it right

If it really was a sample, nothing follows. There is no adjustment that would have helped and no lesson worth extracting. That is a real category and it is fine to land in it.

If it was not, then something about the preparation invited it, and that is information about how the next one will go. The whole reason to be strict here is that the strictness is what separates the two, and only one of them tells you anything.

What the strictness is not

This is not an argument that everything is somebody fault. Genuine randomness exists, it is not rare in absolute terms, and a person can do everything sensibly and still be the one pulled out of the pile.

The argument is narrower. Given that both explanations feel identical from where you are standing, and given that one of them requires no change and the other requires work, the one requiring no change will be chosen more often than it is true. Correcting for that bias is the only reason to be strict at all.

This is the last page in the smallest section on this site, and the section is small on purpose. Most of what gets filed as bad luck belongs to somebody, and more often than people expect it belongs to the reader.

• Back to the cases with no author
• When preparation was the cause
• The attribution rule