Plural Vote

Track Record and Press

Plural Vote was founded by a Columbia University in the City of New York graduate in Political Science and Statistics.

As seen on Wikipedia in a pollster comparison, Plural Vote’s 2017 survey of the Virginia gubernatorial race was more accurate than Public Policy Polling (A-minus-rated by Nate Silver’s ranking), Quinnipiac University (A-minus-rated by Nate Silver’s ranking), Monmouth University (A-rated by Nate Silver’s ranking), Trafalgar Group (B-plus-rated by Nate Silver’s ranking), IMGE Insights, and Rasmussen Reports (B-rated by Nate Silver’s ranking). Plural Vote’s outcome of 44.97% closely mirrored the 46% result for Ed Gillepsie (R).

In terms of presidential races, Plural Vote’s forecast accurately predicted the 2020 electoral vote with 309 EVs for Biden, who ultimately won 306. No other forecast was as precise in projecting the presidency in 2020.

Plural Vote’s founder Sean Le Van has worked on an electoral forecast with Robert S. Erikson and served as TA for Andrew Gelman of The Economist, while a student at Columbia University.

Plural Vote is the only election forecaster to use a particular unique approach, which is to combine polling with Google Search Trends data. In addition, most forecasters do not use artificial intelligence (AI) technologies, while Plural Vote leveraging advanced machine learning techniques like Ridge Regression for enhanced accuracy.

Our platform emphasizes transparency. Below is showcased detailed state-by-state predictions, demonstrating Plural Vote’s low Mean Average Errors, especially in swing states.

Plural Vote’s unique methodology has attracted coverage from notable outlets like Newsweek and The Week, affirming its reputation as a precise and creative forecasting tool in close elections.