Against weatherson on how to frame a decision problem

In "Knowledge, Bets, and Interests," Brian Weatherson makes a suggestion for how to frame a decision problem. He argues that "the states we can 'leave off' a decision table are the states that the agent knows not to obtain." I present and defend an example that shows that Weatherson's principle is false. Weatherson is correct to think that some intuitively rational decisions wouldn't be rational if states the agent knows not to obtain were not omitted from the outcomes in the decision problem. This, however, is not true of every rational decision. Weatherson's principle for how to frame a decision problem is open to counterexample.