Accept losses to reduce them

one

Recommend to everyoneoneThe book, "Saying No to Pseudo-Psychology", hasoneThis is a LOW book title andoneThis book should not be called "No to Pseudoscience," but "No to Pseudoscience," because it doesn't talk much about psychology, but a lot about the philosophy of science, as if the content of psychology is just an example to discuss the philosophy of scienceoneTherefore, after reading this book, it will be of great help to understand and apply scientific thinking.

The main inspiration for the issues discussed in this article comes from Chapter 11 of the book, in which the author's title is "Accepting Errors to Reduce Errors".

two

What issue do I want to discuss?

We know that the world is chaotic, and the market trend acts as the firsttwo"Chaos-like system", but also chaos in chaos, then, the problem is, inoneIn an uncertain market, why should I follow a fixed process when it is impossible for me to find a trading system where all orders are profitable?oneTrading on a trading system where some orders will lose money, rather than trading according to your own subjective judgment or sense of the market?

Or inaccurately simplifyoneNext, in a random market, why should we deal with it regularly rather than randomly?

three

In Chapter 11 of "Saying No to Pseudo-Psychology," Keith Stanovich talks about a classic experiment that has been studied repeatedly for decades: the subject satoneredoneIn front of the two blue lamps, the two lamps are randomly lit in sequence according to the probability of 70% red and 30% blue. The experimenter asked them to predictoneWhich light is on next time?

In the early stages of the experiment, the subjects will make random guesses, but as the experiment continues, the subjects will find that the probability of a red light is higher, at which point the meaningful test will actually begin.

If we follow a rational judgment, we should always guess that the red light is on, which would give us a 70% probability of being correct. However, after all, it is not 100% probability, so the subjects are not satisfied. They will keep looking for various patterns that do not exist, and guess according to the pattern, in order to get everyoneGuess right every time. For example, if it is "red, blue, red basket red", go downoneThe next time they guessed blue, the probability that they would be correct dropped sharply to 58%, 12% lower than the 70% probability that they could easily get.

58%, where did that number come from?

(On the premise of randomness and the absence of any pattern) we consideroneIf you guess 100 times, the subject will guess the red light 70 times, and there will be a 70% correct probability, that is, correct 49 times; at the same time, the subject will guess the blue light 30 times, and there will be a 30% correct probability, that is, correct 9 times, plusoneThis is correct 58 times.

From this, we seeoneA curious phenomenon: the pursuit of eachoneGetting it right every time actually decreases our chances of getting it right, and if we accept that we can't get it right every time and take the initiative to just guess the red light, it actually increases our chances of getting it right.

As you can see, that's why we do regular systematic trading in random markets.

If we follow the trading system strictly, we can achieve the results that the trading system should achieve; if we adjust the trading system smartly during the day, our results will be reduced. What if we trade completely independently and reject the thinking of the system's trading? Then the results will be even lower. For example, if we casually guess the red and blue lights, say the number, 50 times the red light and 50 times the blue light, then our probability will be reduced to 50% again.

four

Of course, for the aboveoneThe content of the section, we have a few questions:

oneThe market is not random. We can use theoretical and empirical methods to prove that the market is trending. Therefore, we can find effective patterns in the market. Therefore, the results of red and blue light experiments cannot be used for trading.

In this regard, we should first say that as strong advocates and practitioners of technical analysis programming and trend following, we certainly recognize the existence of trends. However, we need to clarify what is random. Random does not mean that we cannot predict accurately, but that we do not have the ability to predict accurately.

For example, if we can accurately read the angle and force of the toss, the air resistance, etc., we can accurately determine whichoneFace up, but it is impossible for us to read and calculate it clearly in such a short time, so we treat the coin toss as a random event.

For the price movement, it is of course regular (in fact, I don't like the word "regular" very much, and I will elaborate on it later). The rules of the trading system we design are based on the rules. Although it is impossible to make profits every time (i.e., single randomness), we can be sure that as long as we execute firmly, we will definitely make profits in the end (i.e., overall stability). However, outside this system, it is completely random (i.e., single and overall instability), and there may still be regular parts, but we cannot grasp them. In fact, it would be nice if we could grasp this part, adjust the trading system, and add them (which is also the eternal pursuit of system traders).

Therefore, as far as a single transaction made by the trading system is concerned, whether it can be profitable and how much it can be profitable are completely random, and what we gain is only the overall advantage; the market trends and trading opportunities outside the trading system are beyond our grasp and are also completely random. That is, regardless of the nature of the market (andoneWhat is a word I don't agree with), for us, the market is random, if we blindly screen for opportunities in the trading system, or blindly operate opportunities outside the trading system, our results will be tested with red and blue lightsoneLike this, seeking a higher probability, but achieving a lower probability.

Therefore, my requirements for implementation aresixOne word: do not do wrong, do not miss.

two: If I execute it for a long timeoneA trading system, then, you will have your own feeling about this trading system, when you can make a profit and when you will lose, I know, but this feeling cannot be quantified, so you can't join the trading system, but I guarantee that this kind of screening is feasible.

For this question, I will first quote Wittgenstein'soneIn other words, ifoneAn idea you can't fully express, so just explainoneOne thing is that you haven't considered it yet.

Secondly, I raiseoneAn unrelated example: the lottery isoneIt has been available since then, but the sales volumeoneIt wasn't until the 1970s that New Jersey came outoneA new type of lottery, allowing lottery players to choose their own numbers, has greatly increased lottery sales. Why? Because lottery players have a sense of participation and feel that their lucky numbers, birthdays, anniversaries, are more likely, but in fact? Everyone knows.

Finally, I have to admit that because the disc sensation is completely personalized and difficult to test in the laboratory, I cannot deny the role of disc sensation (of course, precisely because of this)onePoint, I personally completely deny the role of disc sense, and definitely do not allow my team to have this thing), but there isoneThis thing is similar to a disk sense, but it can be easily compared statistically.

Feel.

eachoneAny friend who loves watching basketball should have heard of this wordoneIf a player makes consecutive shots, he will say, Oh, he feels good and is in hot condition, and pass the ball to him; but if he misses consecutive shots, he will say, Oh, he feels cold and is in bad condition. Don't shoot again. This impression is not only shared by fans, but also by professional basketball players.

But unfortunately, this is a wrong impression, and there is no such thing as "feel".

This is very good to count, you just need to count it separately.oneOne player lostoneAfter one ball, after two balls missed, missedthreeAfter the ball and inoneAfter one ball, after two balls, put inthreeAfter the ball, go downoneThe impact of "feel" can be known by the shooting percentage of the second shot. Of course, this isoneA stupid kung fu, but isn't science doing this kind of stupid kung fu?

After psychologist Gilovich and others made statistics, they found that thissixThe number is basically equal (each player shoots according to the fixed practice method, and naturally obtains relatively stable results. Okay, I'm looking for various opportunities to promote the rule-based thinking)!

Forget what it feels like, stop lying to yourself.

threeIf everything you say is correct, then why don't people think so?

In the case of random responses rather than regular responses, there will be surprises. For example, if you only guessed the red light, you didn't guess the surprise of the blue light correctly, but I have 9 such chances. Oh, these 9 times are enough for me to blowoneLifetime. We have seen this situation a lot in our lives.

Of course, the more fundamental answer is that people don't have probabilistic thinking at all!

five

That's why I hate the word "pattern."

When we talk about so-and-so laws, the subtext is that it must be so. But there is nothing inevitable in this world, so they don't take the probabilistic correlations you find to heart, but blindly look for self-righteous so-called laws. For example, when we say "smoking is bad for your health", someone might pop up and say, downstairstwoUncle, I started smoking when I was a teenager.oneTwo packs a day, isn't it still healthy and healthy now? It has reached more than 80. That means that since it is harmful to health, you can't have it.oneA long-lived one.

But can we really find 100% in life?

Maybe we will say: I obviously leave the house at 7:30 and arrive at the company at 9:00. What is the probability of this (I exposed the traffic situation in Beijing)?

On the surface, we go out to work in the morning, which is our personal behavior, but in reality, we eachoneEveryone is a part of the city's transportation systemoneIn part, our ability to get to work smoothly and on time depends not just on ourselves, but also on the smooth operation of the city's transportation system. So no matter how early I leave, I can't be sure I'll be there before 9:00; but conversely, I'm sure I'm more likely to be on time at 7:00 than at 8:00.

This is probabilistic thinking, and our goal is not to guarantee that everyoneNo matter what, the probability of pursuing the whole is constantly increasing.

For example, foroneAn industrial production line, no matter how good it is, always hasoneThe rate of defective products; in turn, weoneThe purpose of all actions is to continuously reduce this rate of defective products.

Forget the "law", pursue overall stability, and continuously increase the probability of the whole. The meaning of human behavior is nothing more than this. And the way to pursue it is to constantly find correlation and even cause and effect in randomness, and incorporate it into the regularized behavior, and the cycle is endless.

six

This article can end here, finallyoneLet's talk about the red and blue light experiment again.

Animals andfiveIn this experiment, children under the age of 10 chose the "red light only" strategy;fivePeople over the age of 12 start looking for patterns that don't exist, which ultimately leads to lower grades.

fiveWhat happened to us at the age of 18?