The immediately challenge then seems to be, can we construct a decision problem where having a mixed-strategy decision theory would cause worse outcomes even if it provides better outcomes in all other cases. That is, can we exploit the mechanism for picking which strategy to use?
The immediately challenge then seems to be, can we construct a decision problem where having a mixed-strategy decision theory would cause worse outcomes even if it provides better outcomes in all other cases. That is, can we exploit the mechanism for picking which strategy to use?