Final polls, models and markets see a "blue tsunami"

By the numbers: Five key metrics help distill the extraordinary consensus taking shape on the eve of the midterms.

🔮 95% and 80%: The probability of Democrats winning the House and Senate, respectively, according to election forecaster Nate Silver, whose model projects an average gain of 25 House seats.

• Silver says Democrats could now withstand a modest polling error and still capture the Senate — a remarkable cushion given the GOP's history of outperforming polls in the Trump era.

🗺️ 7 for 7: Democrats lead in every Senate battleground state surveyed in the final New York Times/Siena polls, including by 11 points in Michigan and seven in Texas.

• Trump carried six of those seven states in 2024, though five of the Democratic leads — Maine (+6), Kansas (+5), Ohio (+5), Alaska (+2) and Iowa (+1) — remain within the polls' margins of error.

🏛️ 240: The number of House seats Democrats could win under the upper end of Cook Political Report's forecast — potentially their largest majority since the first year of Barack Obama's presidency.

• Reaching that mark would require Democrats to sweep all 18 toss-up races. Cook already favors them in 222 districts, four more than needed for a majority.

🎲 73%: The probability of Democrats winning both chambers of Congress, according to Kalshi.

• After correctly favoring Trump in 2024, prediction markets have become a closely watched barometer of political expectations. Nearly $100 million in contracts have traded on Kalshi's congressional control market alone.

📊 +14: Democrats' lead on the generic congressional ballot in the final poll from AtlasIntel, which became the nation's highest-rated pollster after its remarkably accurate 2024 performance.

Reality check: Even the pollsters forecasting a Democratic landslide are haunted by the possibility that they're still missing Trump voters.

• New York Times polling chief Nate Cohn warned that white Democrats were roughly 30% more likely than white Republicans to respond to the final Times/Siena surveys — a slightly wider gap than in 2020, when polls overstated Democratic support.
• Washington Post pollsters raised similar alarms after finding Democrat Adam Hamilton leading Republican Sen. Roger Marshall by nine points in Kansas, where registered Democrats were more than twice as likely as Republicans to participate.

What they're saying: Trump and House Speaker Mike Johnson have accused pollsters of deliberately manufacturing the appearance of a Democratic wave to suppress Republican turnout.

• "They're trying to create this idea of a blue wave and they're using the poll numbers to do it," Johnson said Monday, claiming conservatives distrust polling calls because they believe "it's a psyop."

Between the lines: Prediction markets are sometimes touted as more accurate than polls because bettors have money at stake. But their odds often reflect the same polling data and political sentiment, leaving them vulnerable to the same blind spots.

• In Wisconsin, Kalshi gave state Rep. Francesca Hong a 96% chance of winning the Democratic gubernatorial primary the day before the August election. She lost to Milwaukee County Executive David Crowley.
• In Brazil, Kalshi gave Trump ally Flávio Bolsonaro an 84% chance of defeating President Lula ahead of last week's runoff. Lula won, prompting MAGA influencers to amplify baseless claims that the election was rigged.

The bottom line: Voters will ultimately deliver the only verdict that counts. But on the eve of the election, the polls, forecasts and betting markets are all pointing in one direction.