Some Thoughts on Conditional Forecasts – Lessons from the 2020 Election — LessWrong