Figure I
One massacre, four counts and a gap
Reported death tolls for a single event — Tole, Gimbi district, 18 June 2022 — from four sources. The world's principal conflict database has no event for it in our export; the most committed advocacy source overstates it; the independently corroborating sources converge in between.
ACLEDno event for the massacre in our export
none in our export: the Gimbi entries of 14 June to 14 July 2022 are a military operation, 5 fatalities a day
Human Rights Watchfield investigation
Amnesty Internationalsatellite-corroborated
Amhara Assoc. of America3 Jul update, 455 named; 503 on 23 Jun
05001,0001,551
Undercount (mis-coded)
Independently corroborated
Advocacy count, named list
Advocacy projection
The range in the sources is 400–554: about 400 (HRW) to 554 (AAA, 455 named). ACLED’s “~90” is not a smaller estimate of the same event; it is the event mis-classified as a two-sided “armed clash,” which is how a database can record 4,646 Wollega deaths and still be a floor. Reading the corroborated middle, and naming both the undercount and the overcount, is the whole method of this report.