Status Updates From Думай медленно - предсказыв...

Chapter 8: perpetual beta
Here, again the argument is - you need to try, experience, fail and try again.

Chapter 7: overall, this chapter argues that a good forecast is to be updated overtime small by small. Here the Bayesian principle is mentioned.

After chapter 6: i have no clue what i am reading and this chapter is about

Chapter 6 starts by stating that superforecasters share the trait of being good with number, but they do not need any math to come up with good forecast

Chapter 5:
The chapter starts by stating that it does not just require intelligence and knowledge to be a superforecaster, though it is true that superforecasters possess those qualities.
I found it inspiring to read about Fermi’s class, where a question is then converted to a metaquestion: what must be true so that this … is true?

Chapter 4: superforecasters.
This chapter’s purpose is solely to persuade me that the superforecasters are doing constantly good job over the time and not purely based on luck (because if it is, for a large sample set and several periods, the result will regress to the mean).
Then I think this chapter is too lengthy.

The author forms a study called expert political judgment, which lasts for 25 years and collect + judge forecasters whose livelihood involves economic or political judgments.
The study outcomes show that:
- group of people follows a “big idea” ideology does just the same as a chimp throws a dart.
- other group who collects information makes better judgments.

Chapter 3: Judge the judgment.
Estimate with words are difficult to judge, because they can stretch according to the outcome. How probable is “probable”?