How Friend Groups Work in Crowds

How Do Groups Behave in Crowds?

Aveni (1974) challenged classic crowd theory (LeBon, Park, Blumer), which viewed crowds as atomized collections of individuals who surrender their identity and succumb to a “collective mind.” In an empirical study at the fringes of an Ohio State University football victory celebration, he surveyed 204 participants about their group composition. The result: only 26% were alone. 74% were with at least one friend. Of those in groups, 54% were with one friend, 18% with two, 16% with three, and 12% with four or more.

Additionally: 64% of respondents said they encountered other known people during the event — relatives, coworkers, roommates, school friends. The median number of recognized individuals was 4.6. The crowd was far from anonymous. It was a network of interlocking friend circles.

The Aveni Study

This has three implications. First: classic crowd theory must be revised. Behavior in crowds cannot be explained by individual psychology alone; groups are the primary social unit. Every person carries their friendship bonds into the crowd and acts in constant reference to them.

Second: “anonymity” in crowds is a variable, not a constant. Even unattached individuals behave differently when they know acquaintances are nearby. The fear of being identified dampens disinhibited behavior. This has implications for escalation and de-escalation dynamics.

Friends as a Social Unit

Third: models of collective behavior must be multi-level. They need to account for both individual cost-benefit calculations (Berk 1974) and group dynamics. Decisions in crowds are often made in friend-group micro-dialogues, not by isolated individuals.

The finding applies to other collective events: protests, concerts, stadium visits, dinner events (see dinner group size). Everywhere, people act primarily as members of their friend groups.

Crowds Are Not Masses

Aveni’s study stands against the dominant tradition of LeBon (1969), Park (1972), and Blumer (1946). It aligns with Smelser (1962) and Turner & Killian (1972), who introduced group factors into crowd theories but kept them marginal. Quarantelli and Dynes (1970) documented similar group effects during looting in the racial disturbances of the 1960s. Aveni’s methodological contribution is direct empirical measurement — earlier work relied on observation.

The study has methodological limitations: not a true crowd (the researchers themselves call it a “pre-crowd”), a specific context (a university football celebration), and a demographically homogeneous sample (young students). Whether the findings transfer to more heterogeneous crowds (religious gatherings, political demonstrations) remains open. Later research (Reicher, Drury) confirmed Aveni’s finding but extended it with social identity theory: even strangers in crowds can activate shared social identities (“we the fans”) and then act as a group without personal friendships existing.

Implications for Events

The current state of research on this aspect is summarized below.

Frequently Asked Questions

How do friends behave in crowds?

This has three implications. First: classic crowd theory must be revised. Behavior in crowds cannot be explained by individual psychology alone; groups are the primary social unit.

Are crowds really faceless?

Second: “anonymity” in crowds is a variable, not a constant. Even unattached individuals behave differently when they know acquaintances are nearby. The fear of being identified dampens disinhibited behavior.

Why do groups stay together?

Aveni (1974) challenged classic crowd theory (LeBon, Park, Blumer), which viewed crowds as atomized collections of individuals who surrender their identity and succumb to a “collective mind.”

What does this mean for event planning?

Additionally: 64% of respondents said they encountered other known people during the event — relatives, coworkers, roommates, school friends. The median number of recognized individuals was 4.6. The crowd was far from anonymous.