Nate Silver: "Wizard" in the Big Data age

The scientific community has to play like Elon Musk. Cool idol? Of course, except that President Obama wants to "play" with him, I'm afraid every presidential candidate in the future will want to "play" with him, a superstar in the current American political circle, Nate Silver, who is known as "the perfect" in the eyes of the American public. Prophecy, "His prophecy" is called the Oracle of Campaign predictions.

The man had single-handedly beaten all political journalists, political party media advisers and politicians, and had a good view of the number of nerds, and Americans said that Obama's victory was also a statistic. Let's go back to the 2012 U.S. presidential election, Mr Obama and Mr. Romney were generally considered close, and commentators were unable to predict which side would win, but silver the "real" situation by pointing out that the two were not in the same position, and on polling day he succeeded in predicting that Obama would have 90.9% opportunity to get the majority of the votes, and finally he was right about the results of the polls in 50 states. In fact, in the 2008 presidential election he also predicted the final results, the United States 50 states voted in the results of his forecast of 49.

The analysis of Silver's choice was considered by the extremely subtle American political commentary as an unprecedented level, but because he uses Bayesian theory, known as "Witchcraft Statistics," by academics, the frequency school and some conservative statistical scientists are questioning, as well as accusations from political parties and the media, Because Silver's predictions are like military-level and accurate personal-style reports, the entire political and media circles will be "sitting ducks" for every election that follows.

"Big data is a very important tool, not a game"

In the United States, there are countless statisticians, the reason why silver can become a United States data Superman, because he at the point of the election to prove that most of the political commentators are "useless", and his blog on the New York Times website to allow many companies to see the large data application of the authenticity and large data visualization , industry-like hope.

Silve, who soon published his debut "Signal and Murmur: The Art and Science of prediction," after the 2008 election, and Victor Mayer, a data expert at Oxford University, praised the book, "The Big Data revolution will change the way we live, work and think, We don't have to worry about how much the data is relevant to us, the big data will let us see the meaning of "hypotheses", and in the future it will be an important factor in our decisions, because it has reshaped the world before us.

In fact, Silver also began to worry: his reputation will affect the future selection, and will lose the edge of the onlooker. He does not want people to regard him as the Scientific "Wizard of divination", and does not want people to be superstitious in these predictions, "prediction is a rigorous science, not a game, we have to measure whether our subjective reality can match the objective world." ”

From the large data practicality, large data can help us to solve such as disaster forecasts, financial forecasts and other practical problems, but Silver also pointed out that the data is a trap, people often have false to the data of the illusion of danger. "Probabilistic thinking or absolute thinking in the prediction of the opposite direction, if you can not be honest to look at the data, prediction can become a disaster, ignoring the uncertainty of things often lead to serious consequences." ”

Silver An example, the U.S. National Weather Service predicted in 1997 that the flood level of Grand Forks was 49 feet, that the town's levees were limited to 51-foot floods, did not include positive and negative 9-foot errors through historical data, and that the floods reached 54 feet and Grand Forks suffered a catastrophic flood. Japan's Fudao nuclear reactor is another living specimen, The largest earthquake record in the past 45 years was on the Richter scale of 8, and the Japanese earthquake researcher, Fudao the data, they predicted that the area would not exceed the magnitude of the 9 earthquake on the Richter scale, and then build a nuclear reactor at level 8.5, and that in fact Fudao had had a 9-magnitude quake in the past, but that was ruled out, a prediction that eventually took place in Japan and even the whole The ball brought a historic catastrophe.

"The data itself cannot speak for itself, we often carry out data analysis with biased thoughts and personal interests, and we must know that there is a huge gap between what you receive and what you know." "Silve also pointed out that whether to see their own weaknesses also have a greater impact on the analysis of data, because the understanding of their own weaknesses will take a certain way to counteract its role."

In contrast, he believes that the modelling data from the popular survey is more reliable, because without the media's rendering, the direct arrival of public opinion information is more real, people are unable to keep a high degree of attention to political dynamics. As a result, he tries to avoid dealing with the campaign teams on both sides of the selection analysis, because most of the information is a cacophony of distractions, and most political observers tend to fall into the form of presidential debates and canvassing.

"If you want to succeed, you can look for areas that are relevant to large data and less competitive." ”

35-Year-old Nate Silver, who has neither dropped out of Harvard Yale, nor built a mathematical model to rewrite history from the lab, can even say that one of the most likely to be on the streets of America is better than him.

After earning his degree in economics from the University of Chicago, he went to KPMG as a consultant, and four years later became the most regretful four years of his life, and then he began to indulge in online poker, but the money he earned from his friends was enough to make him quit his job by playing poker for a living. When he was six years old, his hometown baseball team Detroit the champion of the American professional Baseball championship, and at that time Nate Silver started contacting and falling in love with all kinds of baseball stats. Before he turned to the political data forecast, he was a baseball player who predicted the results.

Baseball has ignited Silver's passion for data, and silver has ignited a global passion for the pursuit of big data, entrepreneurs believe that "cloud" or "game" will lead the trend of the times, investors also hope that large data to guide them, the researchers hope to build more scientific and effective model ...

"My book can tell the taxi driver how to work and choose guests, but also to the online dating site a little business advice." But in most cases, we can't handle the messy data, and having more information doesn't mean we can predict it better, which means we're not able to make accurate predictions about everything. "In fact, people can be very good at predicting things like weather, but they are especially bad at predicting stock prices, and terrorist attacks are things that we can predict but tend to overlook," Silver said.

Silver's response is expected to disappoint many, and he advises startups to try to get into some less competitive areas first. "Companies can explore areas where no one has been involved, and if companies have effective data in these areas and can provide the appropriate analytical tools, it may be easier to achieve success." ”

In fact, Sillver, who was one of the few people to discover the potential data orientation of baseball, found that baseball and the presidential election were not competitive areas, and then he created a modeling analytics site with 400,000 dollars won back from online betting, Designed to predict the career prospects of MLB baseball players, he later sold the site to Baseball Prospectus.

"There is a subtle and fuzzy connection between skill and luck." When you win money, is it good or good luck? You never know. "Silver points out that the search for the available data, but no one has ever taken this data to analyze the field, less competition is easier to succeed, the company all the benefits, or all of the competitive advantage, are derived from marginal income.

See here, if you are very envious of silver have this superhuman ability, that has to tell you, do not rule out it is his homosexual tendencies to let him have such a powerful "predictive" ability, "I always feel that I am an outsider, I always have an antisocial view." If you were a lesbian from a young age, or if you grew up in a family that believes in agnosticism and religious beliefs, you don't want to believe in the mainstream of society. ”

(Responsible editor: The good of the Legacy)