Dialectics and Practice of Statistical Trap

When you roll the dice and roll five sixes in a row, many people will think that the chance of six appearing again next time will be reduced, because in our opinion, it is not easy to roll a six itself. Plus it has been thrown five times in a row, the chance of throwing a six next time is naturally lower. In fact, the probability of throwing six is still 1/6, which has been learned in our high school mathematics permutation and combination. The probability that independent events will not affect each other. However, although we know this, we still think that the probability has decreased, and there are countless such things.

When we often see the news of plane crashes and crimes, we feel that the plane crash rate and crime rate are getting higher and higher, that flying is becoming more and more unsafe, and that the world is getting worse and worse. In fact, these are just "illusions" caused by media reports focusing on one aspect. So-and-so products are really rubbish, and so-and-so websites have poor service attitude. Such complaints abound on the Internet, and almost no product or website can escape. Is it really as bad as those people say? Not necessarily, because people often only complain about those unfortunate things, and our eyes can only accommodate these negative comments, and we will be particularly impressed by them. So we forget that there are still many positive comments from the silent majority of users.

Like the media, our memory will be selective, especially when we encounter major events, it is easy to recall some things that we think are "abnormal" before and after the incident, and think that this is a sign. In fact, such behavior often happens, but usually these things are useless to us, so we forget them. This is "Monday morning quarterback". I mentioned the example of earthquake caused by high jump. Because the earthquake was special, I thought that the earthquake was caused by high jump because I had done this kind of special thing in retrospect. If there were no earthquakes, I wouldn't remember what I often do. At the same time, I am obviously not the only one who is also jumping. I was lucky enough to be one of them and happened to meet them. What a coincidence.

People in TV plays often say, "What a coincidence?" I said, what a coincidence. Yes, what a coincidence. We often underestimate the possibility of coincidence. For example, when we meet someone who lives in the same constellation Cygnus, we will sigh what a coincidence, and then feel the fate. Take a class as an example. Suppose there are 60 students in this class, the probability that at least two of them are from the same Amanome is over 99%, 50 students are also 97%, and 40 students are 89%. No encounter is a low probability event. In fact, only 23 students are enough to make the probability reach 50%. It should be noted that this only refers to the probability that two people are from the same Amanome, not the specific date given. A basketball player shoots 20 times and scores at least 4 goals in a row, which is almost 50%. Another example is earthquake prediction. Don't think it is difficult to predict. In fact, each of us can make predictions easily. The probability of getting the right earthquake every year is very high. What a coincidence. In addition, many people are making predictions. You can't touch them, but you can meet them.

The same statistical results, expressed in different ways, will make people feel different. For example, most patients will hesitate if they are told that there is a 10% death probability in an operation. If the patient is told that there is a 90% survival rate, the patient will be more likely to choose to do it.

Imagine a host of a live TV program claiming to have super powers. He said that as long as the audience cooperated and paid attention, he could remotely control the coin and let them toss it up with the same face ten times. Successful viewers will call them and tell them that because there are many TV viewers, assuming there are more than one million, then more than 100,000 viewers have been successfully remotely controlled. It is enough to use thousands or even hundreds of people to confirm. This is a variation of the common tricks used in fortune telling, and it always works. Changing the vest will still fool us, and we have many variants ourselves.

The above are just a small number of statistical liars hidden in us. Because of these liars, we are not only cheated by others, but also by ourselves. Don't lie with numbers. The logic of statistics is everywhere, so you should learn and use the ideas of statistics flexibly.

With regard to statistical data, the author told us to ask five questions: "Who said that", "How did he know that", "What did he miss", "Has anyone changed his concept" and "Is this data meaningful?" If you are excited after reading the data, if you have something to say, hold it back first, then remind yourself to ask these questions, and don't find it false after scolding.