Sky-fall myths are no more common near young craters than anywhere else

Tested 25 October 2026 with Berezkin's catalogue of 958 societies (script: scripts/myth_craters.py). For each sky-catastrophe motif we compared societies within 1,000 km of a young impact with all others, shuffling the motif among societies of the same country so that regional habits cannot fake a signal.

Results: former sky has fallen 4.7 percent near versus 3.0 percent elsewhere (p 0.12); world destroyed by fire 14.9 versus 14.2 (p 0.17); patched sky 0.7 versus 0.6 (p 0.35); objects and stones revolt 0.7 versus 4.2 (p 0.73). Old craters, our control, show nothing either. None is a real signal.

The check that mattered: the unrelated earth-diver creation story looked twice as common near craters on a raw count (27.7 versus 14.1 percent) simply because it is common in Siberia and North America, where several craters are. Shuffling within countries removed it (p 0.94). Without that control we would have 'discovered' a false link.

What this does and does not say: the catalogue records widespread story types, not local place-legends. The known crater stories (Henbury, Campo del Cielo) are local legends tied to one spot, which this catalogue was never built to capture. So the result is: impacts left no broad footprint in world mythology, not that nobody remembered them locally.