Persuasion With Reasons

In well-defined situations where all contingencies are known, signals (transmitted facts) may persuade.

But in high-stake persuasion situations -say a pitch of a new business model idea of a founder to a VC, or the decision of a monetary policy committee on the interest rate- facts may already be on the table and what persuades may rather be theories -reasons- of how they should be interpreted.

So to persuade, advocates may present things that are a priori logically/necessarily true if one follows a line of reasoning, but then they often contradict deeply held beliefs of the people people they want to persuade.

In a paper that is now Revise and Resubmit at AMR, Jens Schmidt and I model this as a game with unawareness: 

Which set of reasons should the advocate use to persuade? When should the receiver trust and resolve the contradiction caused by the contrarian theory she learns in favor of the sender?