If you're talking about how we resolve questions when the answer is (supposedly) known - we try to be pretty careful about what sources we tap and have a growing list of subject matter experts we reach out to to understand what may really be going on. We also have partnerships with IEEE, AAAS, and a few other professional societies who are helping both author forecasting questions and resolve questions. This is all a public/open effort so if anyone wanted to get involved themselves by posting forecast questions in addition to making predictions, they can do so.