In the second-to-last run of the 2026 Election Forecast model (the last run is on Election Day itself), Democrats have hit a new high: They now have a 63% chance of sweeping to the majorities the House and Senate on Tuesday.
That is the highest chance of a Democratic double that the DDHQ model has shown in the entirety of the election cycle. And it comes as every leading indicator — polls and prediction markets mainly — have swung heavily toward Democrats in the final days of the campaign.
It comes as Democrats have hit highs in the DDHQ model in their chances of winning the House and Senate individually.
At the moment, Decision Desk HQ projects an 84% probability that Democrats net the three seats they need to retake the House majority.
And the Senate? Democrats now have a 65% chance of winning the majority.
00:00–02:45 Introductions and how the forecast works: Geoffrey Skelley and Maddisen Morrell explain how an election model combines fundamentals, polling and prediction markets.
02:45–06:45 Democrats hit new highs: An 84% chance of winning the House, 65% chance of winning the Senate and 63% chance of controlling both chambers.
06:45–10:45 Kansas Senate as a case study: How the model weights its inputs—and why prediction markets can move more dramatically than forecasts.
10:45–13:00 Why the forecast has shifted toward Democrats: A flood of late polling, including the latest New York Times/Siena surveys.
13:00–18:15 Breaking down the final Senate polls: Michigan, Texas, Maine, Ohio, Kansas, Alaska and Iowa; undecided voters, outliers and margins of error.
18:15–21:15 What could the model be getting wrong? Republican nonresponse, potential polling error and the question of how much weight prediction markets deserve.
21:15–25:00 Maine Senate: Troy Jackson’s improving position, Susan Collins’ campaign and why her past polling overperformance matters.
25:00–29:15 Why this election is different for Collins: Her electoral history, Maine’s ticket splitters and running under an unpopular Republican president.
29:15–32:30 New Hampshire Senate: Why Chris Pappas is heavily favored over John Sununu—plus a chicken tender detour.
32:30–34:30 New York governor: Why Kathy Hochul’s race looks much less competitive than Republicans once hoped.
34:30–37:30 Wisconsin governor: David Crowley versus Tom Tiffany, the Democratic primary and the state’s closely divided electorate.
37:30–42:15 Can you measure candidate quality? Political “wins above replacement,” incumbency, Rob Sand and Abdul El-Sayed.
42:15–44:30 Nevada governor: Aaron Ford’s improving chances against Joe Lombardo—and whether markets are getting ahead of the evidence.
44:30–46:45 Florida governor: Why Byron Donalds remains favored, but David Jolly has made the race more competitive than expected.
46:45–47:45 Does the model use AI? Madison explains the exclusion of synthetic polling and AI-generated inputs.
47:45–53:15 Vermont governor: Phil Scott’s unusual crossover appeal, signs of a closer race and the challenges of interpreting limited polling.
53:15–55:15 Final forecast recap: Democratic control probabilities and projected seat totals of 241 House seats and 52 Senate seats.
55:15–57:45 Building your election-night dashboard: A tour of DDHQ’s Votes Pro tools, customized race tracking and early-voting data.
57:45–59:37 “The Projection,” the post-election analysis and 2028: Election-night coverage with Chuck Todd, what comes next and who might launch a presidential campaign first.
