Skip to content
Home

FiveThirtyEight: statistical journalism and forecasting

Online news site that applies statistical models to elections, sports, and other topics. Founded by Nate Silver in 2008, it publishes probabilistic forecasts, data-driven reporting, and methodological analyses.

Overview

FiveThirtyEight is a U.S.-based website that applies statistical analysis to public affairs and popular culture, especially elections and sports. Founded by statistician Nate Silver on March 7, 2008, the site's name refers to the 538 electors in the U.S. Electoral College. Its coverage mixes forecasting models, polling aggregation, explanatory articles, and data visualization to help readers interpret noisy information.

Image gallery

2 Images

Approach and methodology

FiveThirtyEight is best known for producing probabilistic forecasts rather than categorical predictions. Its models combine inputs such as public polls, demographic data, economic indicators, and historical patterns. Analysts typically weight polls by recency and quality, incorporate structural "fundamentals," and run simulations (for example, Monte Carlo methods) to estimate distributions of possible outcomes and associated probabilities.

  • Polling aggregation: combines multiple polls and adjusts for bias, sample size, and methodology.
  • Model components: may include state or team fundamentals, turnout assumptions, and economic variables.
  • Uncertainty: forecasts report probabilities and margins to reflect sampling error and model uncertainty.

Notable forecasts and reception

The site gained wide public attention after its 2012 projection, which correctly identified the winner in all 50 states in the final model for the 2012 election. In the 2016 election, FiveThirtyEight's model assigned then-candidate Donald Trump an approximate 28.6% chance to win; Trump ultimately prevailed. That probability was higher than some other outlets' estimates, including analyses from The New York Times and The Economist, and prompted discussion about how to interpret and communicate low-probability but plausible outcomes.

Accuracy, calibration, and criticism

An important part of evaluating forecast systems is calibration—whether events assigned X% probability happen about X% of the time. FiveThirtyEight has published calibration checks and methodological notes to assess how well its probabilities match realized outcomes. Critics sometimes point to high-profile misses or interpretive gaps, while proponents emphasize the site's transparency about assumptions and its explicit treatment of uncertainty.

Content, influence, and uses

Beyond election models, FiveThirtyEight publishes data-driven pieces on sports analytics, economics, science, and culture. Its mix of technical explanation and journalism has influenced newsroom practices around polling and probabilistic reporting, encouraging greater attention to uncertainty and reproducibility. The site also maintains explanatory material and archives to help readers examine model inputs and reasoning.

Further reading

Readers seeking technical detail can consult FiveThirtyEight's methodology write-ups and data pages to review weighting schemes, simulation procedures, and historical performance. Topic-specific articles and profiles linked from the site provide context for individual forecasts and the real-world variables that shape them.

Questions and answers

Q: What is FiveThirtyEight and what does it do?

A: FiveThirtyEight is a website that makes forecasts about elections, sports, and other topics. It also publishes articles about those topics.

Q: Who created FiveThirtyEight and when?

A: FiveThirtyEight was created by statistician Nate Silver on March 7, 2008.

Q: Why is FiveThirtyEight called FiveThirtyEight?

A: FiveThirtyEight's name comes from the number of electors in the United States Electoral College.

Q: What is FiveThirtyEight famous for?

A: FiveThirtyEight is famous for correctly predicting the winner of all 50 states in the 2012 United States presidential election.

Q: Did FiveThirtyEight predict the winner of the 2016 United States presidential election correctly?

A: No, FiveThirtyEight did not predict the winner of the 2016 United States presidential election correctly. They gave Donald Trump a 28.6% chance to win, but he did end up winning.

Q: How did FiveThirtyEight's 2016 election prediction compare to other news outlets?

A: FiveThirtyEight's 2016 election prediction gave Donald Trump a higher chance of winning than other news outlets, such as The New York Times and The Economist.

Q: How accurate are FiveThirtyEight's forecasts?

A: FiveThirtyEight's forecasts are fairly accurate and they have analyzed the calibration of their forecasts to ensure that they are giving an accurate representation of the likelihood of events happening.

Author

AlegsaOnline.com FiveThirtyEight: statistical journalism and forecasting

URL: https://en.alegsaonline.com/art/138390

Share