make clear that researcher can believe in different things while being 'rational'

[aaronpeikert]

And the authors are correct that preregistration can remove uncertainty if all readers of a preregistration 1) agree on the analyses plan, and 2) there are no deviations of the preregistration. However, the authors already say 2 is not always true, as people often need to deviate. And 1 is also not true.