Heads may not be better than a Zhuge Liang

.

In the scale of the problem and the scale of the market continues to expand today, relying on the public power of crowdsourcing is a hot solution.

Heads is better than a Zhuge Liang

The contents of the package are many, including collective wisdom. As the saying goes, heads is better than a zhuge. As early as 1906, versatile British Galton (explorers, eugenics, psychologists, the father of differential psychology, the founder of Physiological Metrology, and the inventor of the scientific cake-cutting Act) have discovered the accuracy of swarm intelligence. On one occasion, he came to a fair and saw a weight-guessing game with 800 people in it-guessing the weight of a slaughtered bull. Afterwards, he collected all the estimates and calculated the average, resulting in 1208 pounds. To Galton's surprise, this is only 10 pounds from the true weight of the cow (1198 pounds), less than 1% of the error!

Collective wisdom can be seen.

The mob and the herd effect

But the masses are not always discerning. We've also heard of the rabble and the herd effect. For example, social news sites that use voting mechanisms for popular content prove that internet users have a herding effect.

Sinan Aral, of New York University, conducted an experiment where he posted articles to an unnamed aggregator and manipulated the "likes" and the number of comments at the beginning of each article. It turns out that articles with "likes" and comments from the beginning have a higher prevalence than none. In other words, groups are susceptible to other people's perceptions.

Researchers have found that if groups are affected in the same way, such as interacting or influenced by an external factor, their views tend to focus on a biased estimate. In that case, the wisdom of the group may be foolish.

Eliminate bias/bias

However, neuroscientists Gabriel Madirolas and Gonzalo De Polavieja of Madrid Cajal Cato, Spain, have found a way to eliminate bias/bias.

They found that some people were more susceptible to additional information from some more assertive people, so the researchers separated the susceptible people from the independent thinkers in two groups. It turns out that independent thinkers are more likely to give a sensible assessment. In other words, the wisdom of the group is not desirable, just the wisdom of the self-confident.

But how can you identify a confident person?

The researchers studied an earlier estimate of experimental data. The experiment was to allow subjects to perform tasks such as estimating the length of the Swiss and Italian borders (734 km), and subjects were divided into groups.

After a task has been performed, the researchers will show groups of other groups ' combined estimates (additional information for others).

The researchers then set up a mathematical model to analyze how the group's people absorbed this extra information. They assume that each person draws the final estimate based on two messages. One is their own independent estimate, and the other is the combined estimate of other groups. Of course, the subjects gave the two different weights.

It turns out that people with large estimates of bias give extra information a lot of weight, and those who are confident have a small or even 0 weighting for extra information.

The power of Confidence

The researchers then compared the behavior of these people to the model, and found out how independent they thought, and then divided them into independent thinkers and biased thinkers. Finally, the researchers adopted a comprehensive estimate of the independent thinkers group, which showed that the estimates were much more accurate.

The results show that the average, median or geometric average may not be able to make good estimates for group prediction, and a better collective intelligence can be formed by taking individual factors into account in social dynamics.

The next question, of course, is how to find self-confidence in the real world and develop the wisdom of these people. We believe that the combination of social science and technology will increasingly accurately portray a person's "portrait" with the development of quantitative ego, virtual and realistic mapping, and large data technology. Know who you are, the above problems can be solved, but it will also give birth to a hint of coolness behind.