vs. a typical

today typical (median) middle 50% of days

included in the typical curve unusually quiet (kept, not in the curve) excluded: closure / holiday sensor glitch (bad readings dropped) day shown above · click a dot to load it

when to go

week at a glance

the 5 pm wave

Door sensor events per 10 minutes on a typical weekday: when people arrive vs. when they leave.

entering leaving

through the year

Average crowd during the weekday 4-8 pm rush, by week. Semesters start busy, fade toward finals, and summer is the off-season; hover for the weekly visit count.

notes

• Counts come from Density door sensors on the weight room's 2 entrances (Annex and Main); "people inside" is entries minus exits. The tally resets at 01:00 every night. Miscounts accumulate: over the last 82 days the count sitting at 00:50, when the room is empty, averaged 0.5 people (2% of days were off by 5 or more). Read evening numbers with a few people of slack. Before averaging, 19 days were set aside as closures or holidays (door entries under a tenth of the same weekday's norm), 114 unusually quiet days (under half the norm: Thanksgiving week, breaks) were kept in the history but left out of the typical curves and averages, and 6 days had sensor glitches (readings above 195) dropped. Hours with no door activity count as closed, not as zero people, so early closings and summer hours don't drag the evening curves down. The dot plot under the comparison chart shows which days are which.
• Capacity is 150. RecWell turns on the virtual line at 95% (≈143 people); the count hit 150 on 103 of the last 365 days. Busiest day: Mon, Sep 15, 2025 with 4,452 entries; a normal weekday sees about 1,953.
• The live number and today's curve are fetched in your browser from the same public api.density.io endpoint the RecWell page uses; the history is refreshed daily and last included 2026-08-30 (built Aug 30, 2026). Raw data: rsf/data.json (10-minute count, entrances and exits per day, Pacific time).
• Inspired by Anthony Ozerov's and Michael Lisano's dashboards.