Suppose the placebo effects exists; if you believe you will get better, then you will indeed get recover.
B(h) -> h
Unfortunately, it only works if you believe you will get better - and it could be hard to see why it'd be rational to believe that in the first place. Fortunately, rationalists have a solution to this problem.
We're scientific sorts of people, so we believe in the placebo effect - that is:
B( B(h) -> h )
and we're also logical sort of people, so we believe lob's theorem
B( B(h) -> h ) -> B(h)
and hence we believe we'll be healed
B(h)
and hence, by the placebo effect,
h
And we're healed!
If it's worth saying, but not worth its own post (even in Discussion), then it goes here.