3 ms·

If the experiment is re-run, but it'd never be noticed if it the results were simply cited and taken as proved.