Nate Silver touted the role of statistics in reporting to an audience of hundreds of journalists at the ONA13 conference on Friday, but he will not be paying much attention to the traffic numbers of his own blog FiveThirtyEight – at least not initially.
Silver, who signed a massive deal with ESPN to expand FiveThirtyEight beyond politics into sports, economics and culture, said analytics will not drive coverage decisions in the first months of the venture’s relaunch.
“I’m not sure we’re going to be spending a ton of time in the first six months checking our traffic reports every day,” Silver said in an interview after the keynote. “We’re engaged in the business of brand-building, and with ESPN, it is a long-term play.”
The guidelines for data-driven journalism Silver presented in his keynote speech stress the idea that data require context. Even data that show a positive outcome most of the time must be viewed in the right perspective.
“If you play Russian roulette, you have an 83 percent chance of surviving. That doesn’t make it a good idea,” he said.
Silver emphasized in an interview that FiveThirtyEight’s initial metrics will also be analyzed with context in mind. The context in this case is a brand-building strategy focused on long-term trust gain.
“The objective should be more long-term focus in terms of how you build a brand,” he said. “Just because you have something easy to quantify, it doesn’t mean that it necessarily gives you the right business strategies.”
Silver also warned the audience against overstating the importance of outliers. In his keynote, he said journalists often overanalyze data points that should simply be ignored.
“For example, if there are 19 polls in a row that come out showing Barack Obama ahead in Minnesota, a 20th poll comes out and it’s an outlier that shows Mitt Romney ahead instead. That poll is likely to get a lot more attention,” Silver said.
Silver is applying the same concept to his business strategy for FiveThirtyEight. True to ESPN style, he used a sports metaphor to explain that he will not be chasing big-hit posts that drive traffic, instead favoring contributors who consistently write articles that do well.
“Someone who maybe kind of hits for a lot of base hits is someone who I think might be better over the long run than someone who has a home run now and then but strikes out a lot and can actually damage your brand on the worst day,” he said.