How estimates are
produced.

Wait times are collected from on-site reports submitted by readers visiting a field office, and from previously observed wait patterns at the same office, day-of-week, and hour. Submissions are anonymous; no contact information is required.

• →Reader-submitted on-site reports
• →Historical observations at the same office, day, and hour
• →Published operating hours and posted closures

Reports are weighted by recency and aggregated by office. A single late report is not allowed to drive an estimate alone; the figure tracks the recent median, with outliers down-weighted.

• →Recency-weighted median across recent reports
• →Outlier suppression to limit single-source skew
• →Per-office aggregation, not regional smoothing

An hourly estimate is derived for each office from its observation history. The figure published as the next-hour estimate is a statistical projection; it is not a guarantee of conditions on the ground.

• →Hourly statistical estimates per office
• →Confidence figures derived from sample density
• →Estimates revised continuously as new reports arrive

The index is published on a single open page. Readers can sort, filter to nearby offices via geolocation, set local alerts, and submit their own wait observations to extend the record.

• →Open public table; no registration required
• →Optional geolocation for nearest-office filtering
• →Reader contributions accepted at any office