Super Forecasting by Dan Gardner and Philip Tetlock

Summary
  1. Great read on how to become a much better forecaster by making a forecast, measuring it, revising it and repeating. Must be a clear and measurable forecast where you can determine your accuracy. Must be open minded, self-critical, careful and curious. Does not come effortlessly and must be very committed to self improvement
Key Takeaways
  1. Forecasting is a real skill that can be cultivated rather easily
  2. Average expert a terrible forecaster but can tell a compelling story with conviction
  3. Our skill at predicting lies in what we’re trying to predict, how far into the future it is and under what circumstances
  4. Best way to improve your forecasting skill – forecast, measure, revise and repeat
  5. Setting a clear goal and finding an appropriate measure is extremely vital for any goal in any field 
  6. What super forecasters thought less important than how they thought.
  7. If have a proven algorithm, use that instead of own intuition. Goal one day is to couple strengths of machines and humans to make best predictions
  8. For any belief, must know what disconfirming evidence would change your mind
  9. Always budget some doubt into your opinions and decisions. Never be blindly confident
  10. Forecasts must have timelines, explicit definitions, easy to judge, define likelihood of X happening, accountability
  11. Calibration and resolution are two facets of good forecasting. Calibration – something happens 40% of the time when you predict that. Resolution – confidence and decisiveness
  12. The average guess of big crowds is often very accurate. The wisdom of large crowds
  13. Dragonfly eye – getting very many perspectives is vital to good forecasting
  14. Post mortems are vital to improve decisions and forecasting
  15. Regression to the mean is a valuable tool to test the role of luck and a good thing to keep in mind
  16. Breaking a large, difficult question into smaller answerable questions is a good way to arrive at a good back of the envelope estimate
  17. Being able to detach self and step back from opinions and estimates allows you to vastly improve your thinking and forecasting (Soros). Also, tweaking the wording of the question helps think about it in a different way
  18. Active open mindedness important. Beliefs are hypothesis to be tested, not treasures to be guarded
  19. Unfortunately most people are attracted and side with confidence over competence. Beware
  20. Do not fall into the trap of thinking that something with a high probability will certainly happen
  21. Super forecasters believe less in fate than most. Probabilistic thinking over deterministic
  22. Must blend inside and outside view, weighing outside view more
  23. Super forecasters incrementally update forecasts often as they get better data and weigh relevance and importance of data in updates. Be cautious of under or overreacting to new information.
  24. Perpetual beta and growth over fixed mindset vital
  25. Effective practice must be accompanied with timely, clear and consistent feedback 
  26. Like in any other field, grit, or the passionate perseverance of a long term goal, is very important to improve forecasting skills
  27. Be careful not to let feelings and emotions  masquerade as rational thoughts
  28. Teams of super forecasters made even better forecasts than a super forecaster working alone
  29. People tend to be givers, marchers or takers and perhaps counterintuitively, givers more often end up on top
  30. If have a good group, can be beneficial to extremize opinions as these people are often too conservative
  31. Leaders must be confident, decisive and deliver a vision. However, this goes against many qualities common in super forecasters. This can be overcome by copying Moltke – being humble because everything is uncertain and must adapt to all new info but once a decision is made, must be committed to it
  32. Intellectually humility and flexibility very important
  33. When in a leadership position, let people know what you want them to do but not how. Let them surprise you with their ingenuity
  34. Being a super forecasters is very difficult and exhausting
  35. Average people fall for trap of scope of insensitivity (not altering decision based on scale)
What I got out of it
  1. Great read on how to improve your forecasting skill – active open mindedness, continual updating (perpetual beta), intellectual humility and flexibility all key

Leave a Reply

Your email address will not be published.