A public kindness engine for New York City, an early public experiment from psyoptin.com. The engine finds and prepares; people bring judgment and action.
Why this page exists. Two decades of published work agree on how a city's mood moves: Saturday reads highest and Tuesday lowest, and rain and heat pull it down. If the reading does not behave that way, the method is wrong, not the city. These checks run from the public archive every time this page is opened, and are printed whichever way they come out. They check version 2, what the site shows; version 1 has its own pre-registered test, which is unchanged.
Not enough data yet. It needs at least 4 Saturdays and 4 Tuesdays of readings; so far 1 and 0.
Not enough data yet. It needs at least 5 rainy and 5 dry days; so far 0 and 1.
Not enough data yet. The heat check needs at least 3 days at 86°F (30°C) or hotter; so far 0. In New York that means next summer.
Not enough data yet. It needs at least 30 borough-days of readings beside 311 counts; so far 0.
Tuesdays sounding brighter than Saturdays over several months, or rainy days sounding brighter than dry ones, would mean the scoring or the normals are wrong. That result would be printed here, as it is, and the method page would say what changed because of it.
From 2026-10-10 the Vibe Report asks "How does it feel where you are right now?" with five faces (great, good, okay, low, rough), a borough and, if someone wants, a neighbourhood from a fixed list. Only the daily count per borough and face is kept, with no account and nothing about the person; each device can answer three times a day. Fixed now, before any taps exist: the test is whether a borough's daily share of great and good taps rises and falls with that borough's version-2 reading on the same day, measured as a correlation across borough-days, each borough against its own average. Nothing is reported before 30 days and at least 20 taps per borough-day; days below that are left out, and the result is published here whichever way it goes. The taps are never mixed into the reading. Counts so far: /api/feel.json.
Written to the studies standard. If a figure here does not match the page it links to, the page is right and this study is wrong: write to us and it gets corrected in place, with a note.