30 Minutes with a Stranger

Alvin Chang turns nearly 1,700 recorded conversations into a scrolling portrait of what happens when strangers are asked to keep talking for half an hour. Animated faces, selected remarks, and mood charts preserve the awkward pauses without losing sight of the people inside the dataset.

What is 30 Minutes with a Stranger?

The page begins with two strangers and a clock, then gradually widens into a crowd of conversations. Scroll time stands in for conversation time: figures appear, remarks bubble up, moods shift, and you can choose individual people to follow. The story does not pretend every pairing becomes instant friendship; some exchanges snag on a bad remark or run out of road. Its larger pattern is gentler. In Chang's analysis, most participants reported feeling better by the end, and the visual rhythm lets that result arrive through accumulated small moments instead of one triumphant chart.

What you can do there

• Scroll through a 30-minute conversation timeline
• Select participants and read conversation excerpts
• Compare self-reported mood across the calls

Why we picked it

This is data storytelling with elbows and nervous laughter. The hand-drawn ASCII faces make a huge corpus feel inhabited, while the minute-by-minute structure gives awkwardness enough room to become rapport. It is patient, specific, and unexpectedly difficult to skim without choosing someone to root for.

How to get the most from it

Scroll slowly enough to notice how the timeline and faces change, then click different participants rather than following only the first pair. Compare the beginning, middle, and end mood views, and keep the outlier conversations in sight when the overall pattern starts looking tidy.

Good to know

The conversations were recorded online in 2020 with U.S.-based adults who opted in and were compensated. The story uses pseudonyms and withholds the original videos for privacy. Its mood result comes from participant self-reports, so read it as evidence from this corpus rather than a guarantee about every encounter.

Who made it, and when?

Alvin Chang is the credited creator or organization. The earliest supported launch date we found is June 13, 2025.