About The Book

Superforecasting details the science behind people who make forecasts better than most other people, while allowing us to learn how to improve our ability to make accurate predictions

Personal Rating: 4.5/5

Short Notes

• Every event has multiple outcomes and it is important to go beyond the simple 0 or 1 outcome we are used to.
• Gather as much information as possible before arriving at any decision and define a probability score for the event. Make sure to keep clear notes on how you arrived at the value for future improvement.
• Apply the first principle of thinking to new situations, asking yourself simple basic questions before searching for data and arriving at a probability score for an outcome.
• Update your score based on the latest data, applying the logic of Baye’s Theorem to arrive at a new score.

Long Notes

• Humans tend to think in terms of “yes,” “no,” or “maybe,” but this three-dial mode is inaccurate as we often take things for granted in two cases. It is important to be clear with forecasts and as numerical as possible, defining a time frame and all aspects that can be interpreted in other ways.
• Developing a more probabilistic approach to any issue is essential. Understand that, in the end, you are just stating the probability of an event happening, and even if the probability is 100%, it is only an indication that, based on all the data collected, the event has always happened and there is still a chance that it may not happen this time.
• When asked to predict, adopt the Fermi-style questioning approach, as with guesstimates. The more fundamental questions that are asked, the higher the chances of an accurate prediction. In other words, start by assessing via first principle thinking. Identify the most fundamental fact that is not questionable and limit assumptions to the bare minimum to move forward.
• When making an estimate, always take an outside view and try to obtain statistics from that outside view. Use the stats collected from the outside view as the base rate and adjust the estimate accordingly.
• It is always important to read as much as possible to update your estimates
  • Work from your base rate as much as possible and avoid being influenced by emotions.
  • Don’t overestimate or underestimate the likelihood of any event.
  • Use Bayes’ Theorem to improve your probability based on the base rate.
  • Aim to find data that contradicts what you believe, as this has a higher likelihood of success.
  • What data can I get to prove that my opposite hypothesis is true?
• Aim for accuracy when calculating the probability.
  • Be as accurate as possible to lower the numerical value
    • 60% is better than 50-50
    • 65% is better than 60% or 70%
    • 64% is better than 65%
  • Try to maintain consistency in your figures to ensure clarity in your ideas.
• Update your estimate incrementally based on new data.
• Don’t become too attached to your initial opinions, but don’t be swayed too much by new information either. It is important to keep making small adjustments to your estimates based on the information.
• It is always important to take feedback on your predictions and then learn why you made mistakes. Analyse why you got it wrong and adopt a growth mindset.
• Keep notes on why you made a certain prediction so that when you analyse, you can be sure of your thought process at the time.