Big data predicts World Cup eight ways

Spain, England two consecutive defeats, the group race is eliminated, not only let some fans sad, so that the loss of lottery players, but also by the way to the many predictions of the World Cup to cry without tears.

The World Cup in the big data after the hot, both private and official, the concept of large data used in the World Cup forecast, but these predictions are true? The following selection of the major domestic and foreign eight World Cup forecasts, their prediction methods for a brief analysis to see who is more accurate.

Baidu analysis of the most traditional

According to the test, this year the National College Entrance Examination Composition Topic 18 Volume 12 Volume's composition direction by the Baidu Big data forecast hit, was dubbed "The god Forecast". Therefore, this time Baidu collects the comprehensive data on the net, then carries on the collation, the analysis, finally through the large-scale machine learning and so on artificial intelligence technology, starts to predict the World Cup.

Baidu Big Data Research institute uses Baidu Big Data comprehensive search in the past 5 years the world 987 teams of 37,000 game data, and with the domestic lottery website Le Choi Network, the European SPDEX index exclusive data provider, such as the company set up a data strategic partnership, the gaming market data into the forecast model, This paper constructs the forecast model of the football match of this "World Cup forecast" product. The model involved 19,972 players and 112 million related data, the reference data include Baidu search data, team base data, player base data, odds market data, etc., the team analyzed not only includes 207 national teams, but also include Europe, South America, Asia and other league clubs and low level team information.

Comment: Baidu used traditional statistical analysis, focus on the recent team and players performance, this prediction is the most technically stable method, but by accident factors (such as weather, injury, referee, etc.) have a greater impact. So far, Baidu has successfully predicted the black Horse of Chile, but omitted Costa Rica.

Deutsche Bank calculates the most ridiculous

Deutsche Bank based on the FIFA rankings of the teams, historical record, player composition and casino odds and other factors, the establishment of a quantitative analysis model, and based on complex calculations to obtain a title probability form, from the title probability table selected the top 10, according to the "rotation cycle", thus excluding the 2014 Brazil, The possibility of Italy and Spain winning the championship, then according to another assumption: the strong teams will return, that is, the World Cup team, the future will inevitably win the World Cup or at least a final. Finally, the England team has 6 players from Liverpool, and it was the biggest player in Liverpool in 1966 and England won the only World Cup in history. At the same time, the Deutsche Bank report's author admits that he is Liverpool's hardcore fans, so the final confirmation that England will win the World Cup title.

Comment: Fortunately, the Deutsche Bank report is not the Chinese team fans!

The Goldman Sachs model is the most mysterious

Goldman Sachs has its own set of evaluation methods (named ELO), the most important of all factors, for the World Cup finals week's 32 national team. ELO is the firm's own dynamic system that constantly updates ratings and rankings based on team performance.

To this end, analysts will collect a number of data, including: The world's national football teams in the history of the database of the scores of the team rankings, the two teams in the past 10 games and 5 games in the number of goals; In the past, the number of teams in the World Cup scored better than usual. In the end, they added the data to a certain weight, and they could tell how many goals each team scored on the other side. In this way, from the team to the final final, each game can expect a number of goals, and finally get a "most average" World Cup full simulation results.

Comments: Investment banks have always been a mysterious model to fool investors, the ELO model is a sophisticated black hole, the key link is not to tell, as accurate or not, only God knows.

Strictly speaking, the above several of the World Cup prediction can not be counted as "Big data analysis", but the traditional statistical analysis, although the data "big", but not integrated into a variety of factors, visible in the professional field or believe the classic theory.

Hawking thought the most entertaining

Hawking collects a lot of data, including history, temperature, altitude of the pitch, and so on, all the data are collected, analyze what you do not know beforehand, perhaps can find some rules. Its principle is not traditional analysis, more is based on the relationship of a prediction. Hawking's 19-page analysis is about how to improve England's chances of winning, but at the end throws a final conclusion that the English fans will be sad: the individual is more bullish on Brazil's title. Hawking believes the England first needs to play at an altitude of 500 metres below the stadium, the temperature rise will reduce the chances of winning, in Brazil 15 o'clock local time is the best time to play. From the team's own point of view, the 433 formation is undoubtedly the rhythm of the championship, and must wear the red shirt. Referring to penalty shootout, Hawking believes that the run-up must be no less than three steps, and if not, the chances of scoring are only 58%. The penalty for the upper corner was 84%, with blond and bald players reaching a higher probability of 84%, the striker scoring more than 80%, and the midfield and defender descending.

Comment: Hawking grandfather in recent years very like to predict, but also predicted that the world will perish after 200 years, this entertainment World Cup, is also a black hole to be stained with the Earth. If the forecast is right, it is suggested that the Nobel Prize that belongs to Hawking should be sent for this reason.

The most tedious sports in Cologne

The results of this World Cup forecast by the German Institute of Cologne based on complex computer simulations: Groll of the Cologne Institute of Physical Education led the research team to calculate 100,000 calculations for their own computer simulations, taking into account the world rankings, lottery odds, market capitalisation, qualifying performances, and possible injuries , tactical, climatic conditions, home advantage factors. They predict that Brazil and Argentina will win the championship, the defending champion Spain is likely to stop the group, from the West Dutch War that thriller 5:1 results, it seems that the German simulation is still reliable.

Comment: Germany's rigorous is famous, and unexpectedly did not predict the German title, for Spain is a language, it is estimated that they are now praying that the final prediction is wrong, otherwise as the physicist Chancellor Angela Merkel will not agree.

Panda predicts premature death

Before the World Cup, according to media reports, the China Research Center for the protection of giant pandas said it would send 1-Shing panda babies to predict the World Cup. In the group stage, the organizer will take out three basket to represent the home team's win, and the panda baby will predict the result by choosing which basket of food to use. By the end of the playoffs, Panda babies will also be able to predict the results by climbing trees and running races. The former is for the panda to climb up the tree with a side of the flag of the trees to predict, the latter is two baby pandas wear two jerseys, through who first run to the destination to predict the results of the game. The "Panda Prediction World Cup" event was canceled after the World Cup. Comments: The original to replace the Octopus Paul's national treasures did not use, the Chinese people or lack of entertainment spirit, take this opportunity to publicize the giant panda, why not, in case if the forecast is right, the giant panda base has become a master Holy Land, but also worry about tourism not fire?

Microsoft believes Excel

Microsoft's Bing big numbers have been successful in predicting Oscars and voting for the election. Microsoft's forecasts take into account past game history, home away, location, turf conditions, weather and "crowd intelligence" and many other factors, and use large amounts of public data-betting markets, polls, social media and other online data-to judge the results of each game with large data analyses. It is said that all of this is done in Excel, we have the right to be a software promotion planning it.

Microsoft: I believe that Excel is omnipotent, but the prediction of football is absolutely impossible, but, people say Oscar, general election are predicted right, or see the results.

Yahoo believes online rumors

Yahoo uses Tumblr data from the light blog site to estimate the strengths of each national team, ultimately calculating the most likely winner is Brazil. The Yahoo Research group's analysis is premised on the value of all the World Cup discussions in Tumblr. To find out which countries will compete against each other, the team gives each team an advantage based on the results of the previous game. For each game, Yahoo uses the probability theory named Poisson distribution to estimate the number of goals each team can score.

Yahoo believes that the current most popular social network data, is said to be able to predict infectious diseases and crime scene, do not know the effectiveness of football?

(it is said that the beauty host wear what jersey, what team loses the ball, and the beauty of the shirt choice is by netizens vote, so, also proved the unpopular too much)

Of course, although many people believe that big data can help us predict the World Cup, there are also unpredictable. Three statistics physicists at the Alamos National Laboratory in the United States have conducted a data analysis of major sporting events and found that in baseball, hockey, basketball, rugby and soccer, football is one of the most suspense, with the most uncertain results, and the high probability of underdog winning over strong teams, Even the use of scientific methods has not been accurately predicted.

To tell you the truth, as a statistical professional, not too sure about football predictions, sports can be predicted, football is no exception, but there are too many factors in football, especially in the World Cup soccer match relatively few, interval period is too long, resulting in small amount of data, there are too many subjective factors (such as referees), Sometimes the predictions of this game are no different from fortune-telling.

If you want to ask, why does anyone always predict the right thing? As one newspaper said, there are countless "Paul" in every World Cup, most of which have disappeared from the media after several previous guesses failed. Bailey is not really a crow's mouth, but he has no follow-up report when he predicts success. This is how the prediction masters are trained!