More specifically, from my model of EY: If some information, rationally interpreted, is (internal to the story) evidence for a hypothesis, then this is good evidence (external to the story) that EY intends this hypothesis to be true. And if some information, interpreted according to a common bias, is (internal to the story) evidence for a hypothesis, then this is good evidence (external to the story) that EY intends this hypothesis to be false. Not only is he not trying to trick us, after all; he's trying to teach us rationality skills. So he can put in red herrings; we just shouldn't fall for them!