How verification works

This archive treats a personal trip as an auditable dataset. Every published claim carries five things: an as-of date, a source document, a reference (page, line, or receipt), a document hash, and a provenance chain from raw document to published figure.

The trust model at a glance

Firsthand visited Source-backed Date captured Last reviewed Evidence available

Five signals travel with every record on this site: it came from a real visit, it traces to a document, it carries the date it was true, it shows when it was last checked, and its evidence is one click away. Anything that cannot clear those bars is labeled or left out.

Status legend

Verified — traced to a primary document in the evidence archive.

Computed — derived from verified inputs; method and inputs shown.

Estimated — contemporaneous operator estimate, labeled as such.

Corrected — superseded; original preserved in the revision trail.

The origin

The underlying trip was the World Race — an 11-month, 11-country volunteer journey I took in 2013–14. The narrative context is part of the story; the data layer stays neutral: costs, routes, venues, dates, documents.

Historical data, stated plainly

These are 2013–14 figures with locked exchange rates as of transaction date. They are published as a historical baseline — the anchor for price-over-time comparisons, not a guide to current prices.

Independent evidence classes

Every era of this archive is corroborated by at least two evidence classes that cannot easily agree by accident: the expense ledger (amounts, locked exchange rates), carrier and booking records (flight numbers, route segments), and camera EXIF metadata (capture timestamps and GPS from 4,000+ originals). The photo archive independently reproduces the itinerary — departure frames geotagged at the airports the ledger says we flew from, month boundaries matching capture-date ranges to the day. When three unrelated record systems tell the same story twelve years later, that is verification, not memory.