204. Nate Silver Says: “Everyone Is Kind of Weird”

Episode Summary

Executive Summary: The interview profiles Nate Silver’s role leading FiveThirtyEight and his broader philosophy of data journalism, prediction, and skepticism toward hype. Silver explains how the site blends statistics, programming, and reporting; why “big data” is often misunderstood; how traffic and fame are measured imperfectly; and why sports, food, and curiosity keep him grounded. He also previews the 2016 election and argues that betting markets can reveal perspective missing from political punditry.

Main Topics: FiveThirtyEight’s mission and newsroom model (Priority: 5/5): Silver defines 538 as data journalism—empirical analysis and scientific method applied to reporting—while emphasizing that programming, statistics, and traditional journalism are all part of the mix. The limits and myths of big data (Priority: 5/5): He argues that big data is often abused as a buzzword, that quantity of data is not a substitute for vetting and inference, and that brute-force correlation hunting can produce false positives. Building and scaling a media company (Priority: 4/5): Silver discusses hiring data-minded journalists, the challenges of creating consistent output, and how FiveThirtyEight has grown from a niche operation into a larger ESPN property. Traffic, metrics, and what success really means (Priority: 4/5): He contrasts short-term pageview spikes with long-term loyalty and argues that not everything valuable in media can be captured by real-time metrics. Sports, gambling, and Silver’s personal interests (Priority: 3/5): The conversation covers his poker background, comfort with sports betting, interest in curling, and love of food—especially burritos—as part of his non-elite, human side. Politics, forecasting, and the 2016 election (Priority: 5/5): Silver gives Hillary Clinton as the plurality favorite while noting that many political stories matter less to bettors than to pundits, and he expects 2016 to be more engaging than the midterms. Identity, public perception, and humility (Priority: 3/5): He pushes back on the idea that he takes himself too seriously, describes fame as fleeting, and emphasizes that adulthood means recognizing everyone is “kind of weird.”

Key Arguments: FiveThirtyEight is best understood as data journalism: statistical analysis and coding are tools for journalism, not replacements for it. Big data is not magic; better data vetting, structured inference, and hypothesis testing matter more than raw volume. Many correlations found through brute-force analysis are false positives, so analysts must avoid overconfidence. Media organizations should balance short-term traffic optimization with long-term audience loyalty and editorial value. Hiring for data journalism requires both technical skills and journalistic judgment; neither is trivially teachable, and teams need collaboration. Betting markets can provide useful perspective on political events because they often weight stories differently than the pundit class. Silver’s comparative lack of reverence for politics may help him analyze it more critically than political insiders do.

Data Points: FiveThirtyEight staff size: 25 or 30 people - Silver describes the newsroom as growing slowly and sustainably at ESPN. 2014 launch date: March 17, 2014 - The date FiveThirtyEight first published after moving to ESPN. Traffic growth: about 2 times year over year - Silver says FiveThirtyEight’s traffic had roughly doubled. Median-month traffic vs. Times era: 5 or 6 times more visitors - He says FiveThirtyEight now gets far more regular traffic than it did at the New York Times. Flight dataset size: about 6 million flights - Used as an example of a large dataset that still runs on a laptop in 10–12 minutes. Processing time: about 10 or 12 minutes - The airline interactive is said to be computable on a laptop in this time. 2016 election horizon: more than 2 years out - Dubner asks Silver for an early presidential forecast. Probability framing: not more than a 50 percent chance for any one candidate - Silver says no candidate is clearly above 50% at that stage. Favorite early-life counting feat: to like 2000 - He recounts counting as a child in nursery school/pre-K. Bowling high score: about 180 - Silver gives his best bowling score. Levitt bowling score: 222 - Silver recalls Steven Levitt bowling a near-perfect game. Beer-and-bowling peak: 1.8 beers - Silver jokingly estimates his peak bowling performance after some beer.

Pivotal Quotes: "“we think it's empirical analysis, quote unquote, the scientific method as applied to journalism.”" — Nate Silver: Silver explains what FiveThirtyEight does and how he defines its editorial identity. "“No amount of data is a substitute for scientific inference and hypothesis testing and kind of structured analysis of a system.”" — Nate Silver: He cautions against treating big data as a magic bullet. "“One of the most profound lessons to me about adulthood is that everyone's kind of weird.”" — Nate Silver: Silver reflects on personality, fame, and what he has learned with age.

Implications: Listeners should see data as a tool, not an oracle. For journalism and politics, Silver’s view favors skepticism, transparency, and long-term measurement over hype, while suggesting that markets and analytics can sharpen judgment if used carefully.