What I read today is Chapter 18 of "Thinking, Fast and Slow " (How to make intuitive predictions more appropriate and effective?) The cognition learned:
Intuitive prediction cannot regress to the mean, and is biased, and often produces extreme predictions, and we can easily be full of confidence in this extreme.
We need to improve our ability to correct this bias through training and improve our prediction accuracy on major issues.
Intuitive prediction, with bias, always unable to regress the mean
1. Predict
prediction, we are still making predictions all the time. Rui Dalio wrote "Principle 2" by analyzing historical data, finding key factors such as education, economy, and currency to make very rational predictions about the future development of the United States, China and the world.
In addition to this rational prediction, there is another type of prediction that is intuitive prediction , which is what Kahneman takes us to understand in this chapter.
2. Intuitive prediction
He divided intuitive prediction into two types :
① Empiric intuitive prediction: that is, because of the recognition of familiar clues, the brain quickly presents the solution to the current problem.
② Heuristic intuitive prediction: This prediction is influenced by heuristics, and simple questions are usually replaced by difficult questions. Timely evidence is insufficient and decisions cannot be changed. People often have confidence when making intuitive judgments.
We need to know that our predictions are not either rationally or intuitively. But the two interact with each other and work together.
3. Disadvantages of intuitive prediction
① Ignore the basic ratio. We have discussed the article in the previous notes. Because we are always accustomed to searching for causality and explanations, we always turn a blind eye to the basic ratio, which cannot improve causality and other statistics.
② is not sensitive to the pros and cons of information. Our intuitive brain can abandon irrelevant or wrong information, but it cannot make up for the relatively small flaws in the evidence.
Therefore, once our intuitive brain discovers some connection to the evidence, the principle of seeing as reality will take effect, and our associative memory will quickly and automatically use available information to compile the most appropriate story. As for how we predict the actual results, we don’t have any feelings.
The process of associative memory: inspires associative memory by information and problems, then feedback itself, and finally select the most coherent and reasonable solution.
② cannot regress the mean. Our intuitive prediction is affected by both substitution and intensity matching, resulting in the prediction being completely unable to regress the mean. That is to say, there is bias in our predictions.
People predict as required is that prediction will always be replaced with estimate for the described question, without realizing that the question they answered was not the question asked.
For example, we will predict that every golfer scores on the second day of the second day and the first day of the first day. In this process, we will replace the second day of the game to establish the first line. The actual situation is that the player who performs extremely well on the first day will most likely return to his daily average level the next day.
Correction of intuitive prediction
Intuitive prediction is biased because it does not have a regression line, so it needs to be corrected. The correction process requires the intellectual brain to participate directly.
1. Four steps to correct intuitive prediction
① Find the baseline . That is, the predictions you make based on not knowing other relevant information, which is the basic ratio introduced in our previous notes.
② Make your own intuitive predictions . That is, predictions made through intuition based on other relevant evidence information.
③ Make an estimate of the correlation between evidence information and the baseline .That is, it makes you move away from the baseline and leans towards your intuition, but the distance you can leave depends on your estimate of relevance.
④ Make corrections to your intuitive predictions . That is, based on the correlation degree (we can set it to a parameter between 0-1), the correlation degree of the corresponding proportion on the reference line is added. That is, the corrected intuitive prediction = baseline × (1+ correlation degree).
Through this step, although the predictions we made in the end were still influenced by intuition, this influence is much smaller.
For example, in this example in the book, Julie is now a 4th grade student at State University. She can read smoothly at the age of 4. What is her average score?
Step 1: We first find the baseline based on the statistics. The book mentions: People who are familiar with the American education system can quickly get a number, and this number is usually around 3.7 or 3.8.
Step 2: We make intuitive judgments based on the information "She can read smoothly when she is 4 years old".
Because of the principle that our intuitive brain sees it as real. When we see that she can read smoothly at the age of 4, we are sure that her reading ability is very good. At the same time, we will associate the inspiration effect of the intuitive brain and imagine that her reading ability has always been getting stronger and stronger. Then our intuitive brain replaces it, replacing the problem of average grades with the problem of reading ability, and making high predictions. For example, 4.
Step 3: Mobilize the rational brain, let’s analyze the correlation between the evidence information "she can read smoothly at the age of 4" and the current senior average score.
4 years old and senior year, there will be a lot of things going through between them, and I think the correlation between them will not exceed 0.02.
Step 4: Correct intuitive prediction. Corrected prediction = 3.75×(1+0.02)) = 3.825.
There are still many intuitive factors in this process above, such as how to define the value of the correlation degree, etc. In addition, it often takes a lot of effort to find relevant references, estimate benchmark predictions, or evaluate the quality of evidence.
So, We should only make such correction efforts when the risks are high and you are particularly eager to avoid making mistakes.
2. Correcting your intuition may make your life complicated.
One feature of unbiased prediction is that people are allowed to predict rare or extreme events only when the information is very effective.
If you expect to make appropriate and effective predictions, your prediction results will never be too outrageous or deviate too much from the average.
If your predictions are not biased, you will never have a "pleasant experience" of extreme events.
3. Black swan is unpredictable .
means that if you want to obtain excess returns, you cannot rely on predictions to complete it, whether it is the intuitive prediction of the intuitive brain or the statistical logic prediction of the rational brain.
Yesterday I also saw a short video, an angel investor said that his investment principle is to only invest in things that you cannot understand and understand. Ten years ago, a man recommended him a matter of buying digital land online, and he couldn't understand at all, so he couldn't make predictions.
He invested 1 million according to his own principles. Ten years later, he heard that JJ Lin purchased land in the virtual world, and the company that sold his land was the company that he angel invested in back then. The 1 million that year had turned into US$2.4 billion.
4. Taleb barbell strategy (you can not read this part)
In addition to daily training of my ability and making corrections to intuitive predictions, for bold goose events, you can also refer to Taleb's barbell strategy for decision-making.
barbell strategy: refers to any strategy that consists of two completely different types of solutions and abandons the ambiguous intermediate route.This strategy eliminates adverse factors through a combination of proactive and conservative paranoia, protects oneself from extreme harm, while allowing favorable factors or positive "black swans" to work naturally and eliminates devastating risks.
Features: ① Asymmetry. It often forms a favorable asymmetry; ② Combination of two extreme strategies; ③ No need to pay attention to intermediate factors
For example,
Reading: On the one hand, we read good books, books written by excellent authors such as "Thinking, Fast and Slow", and on the other hand, we can entertain ourselves by watching movies or TV to relax. As for the mediocre and ordinary books in the middle, we don't need to waste time, and we don't even need to draw generals in it.
Fitness: On the one hand, it constantly challenges the limits and stimulates muscle growth. On the other hand, through proper rest and scientific diet, your body can be relaxed and nutritionally supplemented. As for the current clock-in, completing 10,000 or 20,000 steps per day, this continuous and standard exercise method, from the perspective of people around, is currently very mediocre.
