Systematizing Epistemics: Principles for Resolving Forecasts — LessWrong