Formulating the preference as a "deontological" rule of "you shouldn't do things that would lead you to believe that the total amount of murders would increase" is sneaking consequentialism into deontology.

This is not at all clear to me. The Kantian Categorical Imperative is usually seen as a deontological rule, even though it's really a formulation of 'reflective' concerns (viz., 'you should not act as you would not have everyone act', akin to the Silver and Golden Rule) that could be seen as meta-ethical in their own right.