This is an old revision of the document!
Statistical Inevitability is a concept I made up describing the reality that things, good or bad, will happen. Theft, murder, fraud and other malicious activity aren't as much a tragedy as they are a statistical inevitability. Humans are complex beings and asking “why are there bad people out there that would do these things”1) is not a particularly useful starting point to discuss such events. The amount of certainty depends on the topic in question, but, given a large enough population, most things are statistically certain:
- Typewriting Monkey producing a copy of shakespeare: 1 in 4.4 × 10360,783
- Murderers: 0,006% of people are murderers2).
- A person doing a thing they were not meant to do and for which a clearly visible and accessible notice has been posted beforehand: See Banner Blindness on Wikipedia or other relevant issues like this.
Rates
The point is, the talk about individuals is relevant only insofar as that their motives and the factors acting for and against them or their motives need to be analyzed. Shoplifting is a mass phenomenon and efforts against it need to consider 1) why people shoplift and 2) what makes them stop doing it. The factors causing people to “do bad things” affect everyone. If we say that behavior is a product of circumstances then all it takes is for the circumstances to align just right to cause whatever kind of behavior you want to analyze. The circumstances do not align for most people to cause them to commit murder, but just adding more people into the population is enough for circumstances to align, randomly or not, until one of them will eventually happen to align. To compare regions or demographics or other subsets of a population, you will therefore almost always be talking about “rates”, not absolutes. “10 murders” can be much or little depending on whether we are looking at a population of 11 people (now 1), or a population of 3 million. Most population-based rates these days are expressed as X in 100000 people.
Theft
Theft is a particularly interesting phenomenon to talk about in the context of statistical inevitability, because it is a nice visualization of how real world factors end up in the numbers we see in our spreadsheets. I talked in Theft about the three different factors that determine the morality behind actions, and they play out very nicely in this behavior. There are very few things that would incline a person to murder, AND it's seen as highly immoral by most people. In contrast, trying to designate grocery store theft as immoral struggles with argumentative resistance at every step of the way