- Read some expert advice. The best advice on forecasting is still written up in Philipp Tetlock’s book Superforecasting. I like Daniel Kokotajlo’s summary of the most important points. Farnam Street has another one. Read at least one of the summaries.
- Practice. Start forecasting. I recommend either Good Judgment Open or Metaculus. Make many predictions. Read the questions very carefully.
- Get feedback. Share the reasoning behind your predictions with the community on the platforms. Draw lessons from any feedback that you get.
- Learn from others. Take a look at the models and reasoning of the best forecasters in the community. Emulate them.
- Use prediction and betting markets. Platforms like PredictIt.org contain valuable information. Use it where applicable.
- Improve your calibration. Use this calibration app by the Open Philanthropy Project. Practice more playfully with the Confident board game. You can find more tools I haven’t tested here.
- Learn how to apply Bayes’ rule. Listen to Spencer Greenberg explaining the basic idea in this podcast episode. Practice with a tool he developed.
I never attended one, but you could also consider signing up for a Superforecasting workshop run by the people behind the Good Judgment Project.
You can also find more resources here.
Comments are closed.