Typicality
The assumption that our observation is not an extreme draw from the relevant population — powerful enough to predict and dangerous enough to mislead.
Lukasz Szramuk ·
Typicality is the move from ‘observers can exist here’ to ‘we should expect to be fairly ordinary among the observers who exist.’ The first statement is a compatibility condition. The second is a probability rule. Cosmologists need some version of it when a theory produces many possible observations: without a rule saying which observations are normal, the theory can accommodate almost anything and predict almost nothing.
A typicality argument usually has three pieces: an ensemble of possible worlds or histories, a measure assigning weights to them, and a reference class saying which observers count as comparable to us. Change any piece and the answer can move. In the Doomsday Argument, treating your birth rank as typical among all humans favors histories with fewer future humans. In an eternal multiverse, treating us as typical observers can make Boltzmann brains fatal to a theory because the theory predicts that almost every observer is a fluctuation.
Typical does not mean average in every property. A typical human need not have average height, income, language, and location simultaneously; almost nobody does. It means the observation should not occupy a deliberately exceptional region of the probability distribution chosen for the question. This is why typicality can test a model, but only after the distribution and comparison class are stated rather than smuggled in.
Break the Universe uses typicality as a storytelling convention, not a theorem. A surviving universe is treated as observer-friendly, and the Doomsday toy treats a birth rank as a random draw. Both pages point to where that assumption enters, because the result is only as honest as the population it samples.