The sum of individually rational choices does not reliably produce an outcome anyone chose or prefers, because each person responds to an environment that everyone else's choices create. Understanding a social pattern therefore requires modeling the interaction between motives and the system that aggregates them, not just reading the pattern as a scaled-up version of individual wants.
Mild individual preferences about the composition of one's surroundings, such as not wanting to be a small minority, can cascade into near-total segregation even when no one wants a segregated outcome. Each departure shifts the ratio for those remaining, triggering further departures, so a stable integrated mix is hard to sustain.
Many behaviors depend on how many others are already doing them, so each person acts only once participation crosses a personal threshold. If enough low-threshold actors go first, they pull in higher-threshold actors in a chain; below that starting level the activity collapses, producing multiple possible equilibria from identical populations.
Situations where each person picks one of two options, and the payoff of a choice depends on how many others make the same choice, generate curves where individual optimization lands the group at an outcome worse than an available alternative. The gap appears because each chooser counts only private payoff and ignores the effect of their choice on everyone else.
To explain a social pattern, trace what each participant is actually responding to: the costs, numbers, and others' behavior as they personally experience them, not the overall result an outsider sees. The visible macro-pattern is often an equilibrium of these micro-responses rather than evidence of anyone's intent.
Audiences often leave the front rows empty and cluster toward the back, an arrangement no one designed and the speaker dislikes. Each person seats themselves relative to where others already sit, so individually reasonable placement produces a collectively suboptimal distribution that persists.