Starting with the Iowa caucus on February the 3rd, 2020, American voters will begin casting votes for the Democratic nominee they wish to compete President Trump. In our statistical model, we forecast which Democratic candidate will win the largest number of votes. We factor in opinion polls and betting odds in order to form a comprehensive forecast for nine of the most high-profile Democrats that have declared their candidacy for president. Our forecast automatically updates on a real-time basis as betting odds and polls change.
Last updated: Wed August 21 2019
Probability of winning largest popular vote share
Joe Biden 57.2% chance
IN PAST day week month
Bernie Sanders 19% chance
IN PAST day week month
Harris 5.2% chance
Warren 7.4% chance
Buttigieg 3.2% chance
O'Rourke 2.1% chance
Booker 2.1% chance
Yang 1.9% chance
Gabbard 1.8% chance
PROBABILITY OF WIN
EXPECTED WIN pct. %
% VOTE SHARE
80% OUTCOME RANGE
27.9% TO 55.7%
Statewide polling average: 34.7 %• Current national betting odds: 29 %LAST UPDATED TODAY
We currently estimate that Joe Biden will win the popular vote with an 19-point advantage over Bernie Sanders. Joe Biden has a 57.2% chance of winning the most votes in the 2020 Democratic primary and Bernie Sanders has a 19% chance.
67% vote share330
Current Biden polling advantage 19%
New Year 2020 →
22.8% Bernie Sanders41.8% Joe Biden
Wed August 21 2019
How accurate are our predictions early on?+−
We weight candidates' vote share prediction to be 90% of opinion poll averages and 10% of betting odds in our national forecast. For our primary forecasts, we weight betting odds to be 5%. We then model the probabilities of our predictions using logistic distribution and past polling error. In early October 2015, our predictions would have shown Donald Trump winning the most votes of any GOP candidate with a 27.8% share; he ultimately won with 45.0%. Hillary Clinton was predicted to win with 62.2% of the vote and Bernie Sanders 36.4%; she ultimately won with 55.2% and he with 43.1%.
Our predictions (using data available in October 2015) of 2016 primary candidates' ultimate vote share were, on average, within 8.8 points of their actual result. Each candidate's forecast correlated with the final result by r = 84.4% and r² = 71.2%, proving predictive even in early stages of the primary process.