For each year, we take the share of that year's reports made up by each shape (not the raw count) — this controls for the huge overall growth in reporting since the 1990s, which would otherwise make almost every shape look "on the rise" by coincidence. We then correlate that yearly share against the year itself (Pearson's r) for the ten most-reported shapes, using years from 1950 onward with at least 5 reports.

Raw abduction counts are dominated by the U.S. simply because most of this dataset is American. Dividing each country's abduction count by its population (reports per million people) controls for that — though it introduces its own distortion: a country with just 1-2 abduction reports and a small population can rank highly here even though that's barely any signal at all. Population figures are rough, current-era approximations, not tied to any specific year.

For each of the 10 countries above, the top 3 cities by abduction report count.

Click any point on the map — a sighting, a PURSUE record, anything — and choose "View full case file" in its popup to see the complete details here.

Drag the slider to jump to a year. The event history builds up as you move forward (events don't disappear once passed) and the shape ranking below is a running total from the earliest reports up to the selected year, each shown as a proportion of that cumulative total, always sorted most-frequent-first.