What it ranks on
Of the publishers carrying a story whose article has actually been read, how many name no news agency in the body.
The newsroom counts separately owned outlets. That is a real measurement and it still cannot tell eighteen publishers from AFP printed eighteen times. Both read as 18.
One window had the Riyadh airport story and the Ed Sheeran story both showing 18 outlets. Eleven of the fourteen Riyadh articles that had been read named AFP or Reuters, against six of fourteen on the other. Same number in the newsroom. Not remotely the same thing.
Read is the operative word
Every row says how many publishers were read and how many have not been, and the unread ones never fold into the score.
An early version reported “4 of 4 name no agency” off two articles. That is a confident-sounding number resting on nothing. A count without its confidence is a different count.
A story reaches this list with at least three independent publishers and at least two articles read, which is why it is shorter than the newsroom: 60 rows from a pool of 61, capped at 60.
“First seen here” is not “broke it”
Each row names the first outlet this site saw with the story. A claim about this index, and nothing more.
Rank by publication timestamp alone and you get Daily Trust breaking an Arsenal match report and the Economic Times breaking a White House announcement. Both were agency pickups that declared an early time. Publisher timestamps cannot carry an origin claim, so this page does not ask them to.
The movement figure
+5 in 36 min is independent publishers gained between two runs, from the record kept each hour.
The newsroom has no sense of time at all. A story that went from two outlets to fourteen in an hour and one that has sat at fourteen since yesterday look identical on it.
What it cannot tell you
An outlet naming no agency is not thereby doing original reporting. It may have rewritten a press release, followed another outlet without crediting it, or just not said where the material came from.
What this measures is the absence of a shared named source. Weaker than independence, and more honest.
And as everywhere else here: not a truth score. Ten publishers working separately can all be wrong.
The rest of it
How stories are grouped, which publications are read, how often and how politely: how this is measured.