Fieldwork in Horizon Worlds, Meta’s social VR platform

Meta released Horizon Worlds to the public in December 2021. Our course team asked how its design shaped community: where people gathered, how they handled trolls and harassment, and how its creators learned to build.

From February 1 to March 29, 2022, each of us visited different worlds and events and wrote field notes after most visits. I joined the weekly LGBTQ+ meetup and the Horizon World Tours creator events, and I interviewed three of our six participants.

I spent my first few weeks in the meetup listening before asking anyone for an interview, so that recruiting would not intrude on a support space.

Recruiting was difficult. Most encounters were brief, with no expectation of meeting again. Casual visitors rarely agreed to an interview, while world creators often did, so our interviews lean toward people who build on the platform.

We analyzed the notes and interviews against Amy Jo Kim’s principles of community design, Ray Oldenburg’s description of third places, and Lave and Wenger’s communities of practice.

Oldenburg describes third places, such as cafes and bars, as neutral, accessible, playful places where conversation is the main activity and regulars return. Horizon Worlds as a whole did not fit that description. Some worlds were timed games that expected players to stay, and private worlds were not open to everyone.

Individual worlds came much closer, and their builders shaped the conversation. Builders set how far voices carried: in one shooter game everyone could hear everyone, meditation worlds kept audio short-range, and a comedy club amplified whoever held the microphone.

In my field notes, the Soapstone Comedy Club worked like a late-night bar for adults. Most people gathered at the stage while a smaller group talked about work and their day at the bar. No one pressed me to speak, so I could observe without unmuting.

We wrote a 38-page paper and presented our findings to the class in spring 2022. The study describes the community; it did not propose or test design changes.

Our six interviews lean toward world creators, and fieldwork covered nine weeks of a platform that changed while we studied it. The findings describe Horizon Worlds in early 2022, not social VR in general.

This was my first study of moderation in practice. I later studied misinformation interventions and media provenance at the University of Washington.

The clearest signals were workarounds: creators scripted removal tools, a community blocked trolls one person at a time, and commissions moved off the platform.

These come from nine weeks in early 2022 and interviews that lean toward creators. They are questions a platform team could test, not findings about social VR.