Methodology
How Timepath works
Timepath is the front page of history: curated Timepaths of the world, and of every person, place and organization in it, built on one rule. Nothing appears without published reporting behind it.
Where events come from
Journalists report the news on their own publications — from international wires and newspapers to local outlets. Timepath's automated engine reads that reporting and curates it into events: for the World Timepath, one defining event per day — the occurrence most likely to appear in future history books — and for each subject's Timepath, only the moments that shaped its story.
Every event links the published articles it was built from. You can always click through and read the original reporting — the sources are the point, not a footnote.
How everything links
A single piece of reporting connects four things on Timepath. Follow the chain from the article you can read to the Timepaths it shapes — and note that every publisher and every journalist is a Timepath in its own right.
Published article
The reporting
on the publisher's own site
Event
A verified moment
backed by its cited sources
Subject Timepaths
The story it shapes
the people, places & orgs it affects
Open a source's publisher or reporter anywhere on Timepath to see their Timepath — the full record of what their journalism has shaped. When a reporter or outlet isn't on Timepath yet, it stays plain text until it is.
What the truth score means
Events built by our consensus engine carry a truth score from 0 to 100. It is not a verdict on whether something is true — no algorithm can give you that. It measures something narrower and checkable: how broadly the story was corroborated, and how well the reporting behind it was sourced. Four independent pillars add up to the number.
| Pillar | What it measures | Max |
|---|---|---|
| Editorial consensus | Distinct front-page outlets running the story | 30 |
| Citation & sourcing | Density of external and primary sources cited | 25 |
| Publisher authority | Authority tier of the publishers involved | 25 |
| Author verification | Verified journalist bylines, not desk fallbacks | 20 |
Everything is counted per publisher, never per article. An outlet that runs five follow-ups on one story is one voice, not five — otherwise a single newsroom could manufacture a high score on its own, which is exactly the pattern this score exists to catch. The total then becomes the label on the badge, and the five-bar meter beside it fills in step with the score:
| Score | Label | What it tells you | Meter |
|---|---|---|---|
| 75–100 | High Consensus | Widely corroborated by authoritative outlets with named bylines. | |
| 55–74 | Verified Sourcing | Solidly sourced, but not yet at the breadth of a full consensus. | |
| Below 55 | Developing | Real reporting, still thin — treat it as a story in motion. |
Two rules override the score, because breadth of corroboration is the question a reader is actually asking. A story only one publisher has run is always labelled Single Source, however authoritative that publisher is. And fewer than three distinct publishers can never be called a consensus — the strongest honest label there is Developing, no matter how the pillars add up.
Hover or tap the meter on any event to see its four pillars broken out — the score always shows its working.
Events without a score
Not every event comes from the scoring engine. Our daily world feed is assembled by a different process that doesn't score, and older events predate scoring entirely. Those keep the original badge — "Verified by N sources" — which simply counts the published articles the event cites, with the same meter turning that count into a scale: more independent reporting on the same occurrence means higher consensus. This fallback is permanent, not a migration step. These are its exact thresholds:
| Cited sources | Consensus | Meter |
|---|---|---|
| 1 | Reported | |
| 2 | Corroborated | |
| 3–5 | Strong | |
| 6–9 | High | |
| 10+ | Very high |
Older events are still being enriched with their source data; when an event has no source record yet, its trust badges simply don't show — we never fabricate a number.
What we never do
- ·No user posts. Readers explore, follow and share nobody can write onto a timeline, so there is nothing to moderate and no viral noise.
- ·Journalists don't write on Timepath either. Their work lives on their own publications; Timepath maps it. Verified journalists get an automated portfolio of impact, never a publishing tool.
- ·No event without a citation. If the reporting isn't there, neither is the event. A short Timepath means the story has few defining moments not that we ran out of content.
Don't take our word for it
Every claim on Timepath is checkable: open any event's sources and read the original reporting. As newsrooms publish corrections and follow-ups, our engine re-visits events and their sources on a rolling basis.