Brazil's Thrashing Shakes the Prestige of a Famous American Statistician

07/14/2014 - 09h03

OMG ("Oh my God"), wrote mathematician Nate Silver on Twitter when the Brazilian squad began to crumble in the face of the German team.

The implosion of the team has also shaken 36 year-old Silver's reputation, the most celebrated American statistician, who had predicted a 65% chance for a Brazilian victory, even without Neymar and Thiago Silva.

His website, FiveThirtyEight (538, the number of representatives in the American electoral college), bet in June that Brazil had a 44% chance of being the champion.

His predictions are taken very seriously in the US. Silver famously hit on almost all the results in the 50 states during the presidential elections of 2008 and 2012 - most of the time, on a blog in the "New York Times".

He loves sports and created a model for predicting (and getting right) baseball results.

In football, he used two systems that crisscrossed up to 10,000 simulations among the possibilities of results.

The "Soccer Power Index" (a power index for football), measures the advantage of playing at home to the players performance, defensive, the weight of the distance that teams traveled (and the effect of the time zone), with a greater weight for the latest games.

Silver lives the days of Felipão. He wrote in his blog, hosted and maintained by ESPN, that the 7-1 is "the most shocking result in World Cup history", highly "unlikely." Soon after came the "Oh my God" on Twitter.

The jubilation on the part of the American media is due to criticism that Silver for years has pointed to American TV commentators who "ignore data and facts", Silver said in an interview with Folha last year.

Did the World Cup reveal the limitations of the methodology?

"The statistics were overestimating the quality of the Brazilian team, overstated in fouls, that refs were not taken into account," says Zeynep Tufekci, an associate professor at the University of North Carolina and a researcher at the Center for Research in Technology at Princeton.

She suggests that "no statistical data" should be taken into account.

"If you watch a game with three experts, they will count the number of fouls not called, undeserved penalties and could see that the Brazilians, after Neymar's departure, acted as if someone had died in a catastrophe," compared Zeynep Tufekci .

"It wasn't a switch of one player for another like a Lego piece."

Silver released his first book in 2012, "The signal and The Noise," in which he explained the limitations and errors in the art of making predictions.

A spokesperson said it was the mathematician did not have any time to answer questions from Folha.

Translated by STEVE HUGHES