Nate Silver Doesn’t Like Twitter Data

Photograph by Amy E. Price/Getty Images for SXSW

Nate Silver, Founder & President of fivethirtyeight.com, speaks onstage at the 2013 SXSW Music, Film + Interactive Festival in Austin.

Nate Silver’s credibility may be at an all-time high. After accurately predicting how all 50 states (and the District of Columbia) would tilt in November’s presidential election, Silver’s FiveThirtyEight blog correctly picked Louisville to win the NCAA men’s basketball tournament earlier this month.

The 35-year-old professional prognosticator is now taking his show on the road, speaking to energized crowds about the effectiveness of trying to forecast everything from market and company performance to when a deadly earthquake will strike. He was the keynote speaker yesterday at Lithium Technologies‘ annual conference in San Francisco. Lithium provides software to help businesses communicate and collaborate with their customers, so the audience was filled with Twitter-savvy marketers.

But Silver is somewhat of a social-media skeptic — despite his half-million Twitter followers. He doesn’t like to rely on data from those sites to make predictions. That’s because it’s all still so new and untested, he said in an interview before his speech. Silver singled out this week’s fake Associated Press tweet, which said the White House had been attacked. Stocks tanked right away, before bouncing back moments later.

“Say you’re building an automated high-speed trading strategy, and you’re reacting to news events; how much should you react to one tweet?” asked Silver. “People tend, in general, to overreact to individual blips of information.”

Silver, who previously developed a method for predicting the future success of professional baseball players, does have some new projects under way. He soon will be releasing a study designed to help consumers pick flights based on which airlines are most likely to arrive and depart on time. Much of the existing data grading airlines doesn’t factor in variables like which carriers travel to the most challenging airports, Silver said. For example, JetBlue’s on-time percentage is hurt because its planes frequently fly into and out of New York’s John F. Kennedy International Airport, which has a dearth of runways, and San Francisco International Airport, which is plagued by the city’s notoriously troublesome fog, Silver said.

“You can take relatively simple insights and add value that way,” Silver said. “If you’re a traveler, having more context about how airlines are doing relative to their peers would be helpful.”

Silver isn’t done with politics though. He plans to release estimates for the 2014 midterm elections and the presidential contest two years later. In the meantime, he’s not expecting the U.S. Congress to get much done, because members of both parties are too concerned about winning primary contests, he said. “Defaulting to betting on inaction in Congress is probably not a bad strategy.”

Perhaps Silver’s boldest prediction for 2014 is that the Democrats are unlikely to regain control of the House of Representatives, because the political party of the sitting president typically doesn’t do well in midterms, he said.

But Silver has another less daring pick for 2014 that President Obama, an avid Chicago White Sox fan, would probably appreciate. “The Cubs won’t win the World Series in 2014,” Silver said.

The Cubs haven’t won a Major League Baseball championship since 1908. It makes sense for Silver to play it safe after his fumbled prediction that the San Francisco 49ers would beat the Baltimore Ravens in the Super Bowl this year. Can’t win ‘em all.

What do you think about this article? Comment below!