With three days to go before the 2026 midterm elections, election forecaster Nate Silver has raised the Democrats' chances of winning control of the United States Senate to 79 percent, up from 75 percent, in the latest update to his Silver Bulletin forecasting model. The Nate Silver Senate forecast now says it is no longer honest to describe the battle for the Senate as too close to call, even as Silver himself warns that a normal-sized polling error could still flip the result.

Silver announced the revised projection on The Argument, a YouTube programme, and explained the move on his Substack, pointing to what he called very strong polls for Democrats in individual Senate races over the past couple of days. The update lands at a decisive moment: around 48 million Americans have already voted early, Republicans are bracing for what one senior party strategist described to Politico as a "brutal evening," and control of both chambers of Congress will be decided on Tuesday, November 3.

Key Points

• Nate Silver's FLIPR model now gives Democrats a 79% chance of winning the Senate, up from 75% previously.
• Silver says the Senate race is past the point where it can honestly be called too close to call, but warns leads in Texas, Ohio, Alaska and Maine are a normal polling error away from flipping.
• A planned adjustment on the Alaska race could bring the figure to roughly 77% or 78%; Newsweek reported the model at 77.7%.
• CNN data analyst Harry Enten separately projects Democrats could win 223 House seats, above the 218 needed for a majority.
• Democrats need a net gain of four Senate seats to take control from the Republicans' 53-47 majority, with Vice President JD Vance breaking a 50-50 tie for the GOP.

What Happened

Silver, one of America's best-known election forecasters, told The Argument that his model now put the Democratic chance of capturing the Senate at 79 percent. "That's a little bit higher than prediction markets, but I feel decent about that actually," he said, according to reports of the appearance. On X, he wrote: "Very strong polls for Democrats in individual Senate races over the past couple of days. Polls can be wrong, that's what models/probabilities/markets are for, but I think we're past the point where you can honestly say 'polls show the Senate is too close to call.'"

The revision was driven in part by a recalculation in Montana's three-way Senate race, involving independent Seth Bodnar, Democrat Alani Bankhead and Republican Kurt Alme, where Bodnar had earlier been categorised as an isolated third-party candidate. Silver also noted that a planned adjustment to the Alaska contest between Democratic challenger Mary Peltola and Republican Senator Dan Sullivan could bring the nationwide probability down slightly, to roughly 77 or 78 percent. Newsweek reported the Silver Bulletin's FLIPR model at 77.7 percent, a reminder that the headline number moves with small modelling choices.

The forecast update follows a run of strong Democratic polling. CNN/SSRS polls published on October 29 showed Democrats leading in all four battleground Senate races surveyed: Abdul El-Sayed ahead of Mike Rogers in Michigan, James Talarico ahead of Ken Paxton in Texas, Graham Platner ahead of Susan Collins in Maine, and Patrick Hamilton ahead of Roger Marshall in Kansas. Silver's caution, however, was explicit: in his Substack post he wrote that Democratic leads in Texas, Ohio, Alaska and Maine were "all just a normal-sized polling error away from flipping."

Background: How the Senate Map Works

Republicans currently hold the Senate 53-47, so Democrats need a net gain of four seats to reach 51 and take the majority outright. If the chamber splits 50-50, Republicans keep control because the vice president casts the tie-breaking vote. That arithmetic is why a handful of races matter so much: Democrats must defend their own seats and flip Republican-held ones in states including Texas, Ohio, Maine, Kansas, Alaska and North Carolina, while Republicans are targeting Democratic-held seats in Michigan and Minnesota.

The 2026 midterms are widely seen as a referendum on President Donald Trump's second term, with rising living costs and the protracted United States-Israeli war against Iran shaping the debate. Trump's approval rating has fallen to record lows in recent polling, with a Reuters/Ipsos survey cited this week putting it at 31 percent. Trump has dismissed unfavourable polling as "fake." The president has spent the closing weeks on a 32-day campaign blitz, including a rally in Daytona Beach, Florida, on Friday night, and restored rallies in Kansas and Iowa for Sunday after briefly cancelling them.

Silver's model, known as FLIPR, converts polls, historical patterns and uncertainty into probabilities rather than certainties. A 79 percent chance means that, in the model's simulations, Democrats win the Senate roughly four times out of five — and lose roughly one time in five. Earlier in October, Silver had put the Democratic Senate chance at about 75 percent, and a week before the election he had described an roughly 11-point anti-incumbent swing, against a historical midterm average of about 7.5 points, as the force turning even Ohio, which Trump carried by 11 points, into a contest and Texas into a genuine fight.

Nate Silver Senate Forecast: Why the Model Moved

The move from 75 to 79 percent is modest in itself, but its message is not. Silver is arguing that the direction of the race has become clear even if individual states remain close. Strong state polls, rather than a single national number, drove the change: when several battleground surveys move in the same direction at once, a model becomes more confident even though each individual poll carries a margin of error of four or five points.

That is also the model's vulnerability, and Silver does not hide it. If polls are systematically off by three or four points in the same direction — the kind of correlated error that has bitten forecasters before — most Democratic Senate leads evaporate at once. His warning about Texas, Ohio, Alaska and Maine is therefore the most important sentence in the update: the forecast is a statement about probabilities, not a prediction that the result is settled. Prediction markets, Silver noted, are slightly less bullish on Democrats than his model, a gap he acknowledged rather than dismissed.

Inside Elections, another nonpartisan handicapper, moved in the same direction on Thursday, shifting its ratings in favour of Democrats across nine Senate races, including Georgia, Ohio, Texas, Iowa, Kansas, Nebraska, South Carolina, Louisiana and Mississippi. The convergence of independent analysts matters because it suggests the shift is in the underlying polling data, not in one forecaster's assumptions.

Reactions and Quotes

"Very strong polls for Democrats in individual Senate races over the past couple of days," Silver wrote on X, adding that it was no longer reasonable to call the Senate too close to call. On The Argument, he said of the 79 percent figure: "That's a little bit higher than prediction markets, but I feel decent about that actually."

His caution came in the same breath. "[The leads] are all just a normal-sized polling error away from flipping," he wrote on Substack, a warning aimed at Democrats tempted to treat a forecast as a result. CNN's Harry Enten, looking at the House rather than the Senate, was blunter about the lower chamber: "Look at it now, up like a rocket. We're now talking 223 seats. This is north of the 218 that Democrats need to take back the US House of Representatives," he said on Wednesday.

Republican strategists, speaking anonymously to Politico, did not dispute the direction. One senior Republican strategist described the coming election night as a "brutal evening," warning: "Once the dominoes start falling it becomes hard to stop them." Another senior Republican was quoted as saying: "Any Democrats telling you they aren't winning the Senate [besides] a couple of seats are sandbagging expectations." Trump himself has dismissed unfavourable polling as "fake," and his campaign has pointed to its own turnout operation, including nearly 20 tele-rallies in the final 72 hours, as the counterweight to the polls.

Why It Matters

A Democratic Senate would transform the final two years of Trump's presidency. With the House also projected to flip, Democrats could block the administration's legislative agenda, launch congressional investigations, and control confirmations for judges and executive appointments. A Republican hold, by contrast, would preserve Trump's ability to confirm nominees and protect his agenda even if the House changes hands.

The forecast also matters for what it says about the electorate. An 11-point anti-incumbent swing of the kind Silver described would be larger than the historical midterm average and would signal that voters are punishing the party in power over the cost of living and the Iran war, rather than rewarding it for economic growth figures such as the 3.4 percent annualised third-quarter expansion reported on October 29. Early voting data already points to extraordinary engagement: 47.9 million ballots had been cast by midday on October 30, surpassing the previous midterm record of 45.9 million set in 2022.

For voters, the practical lesson is narrower: forecasts do not vote. A 79 percent chance still leaves a one-in-five path for Republicans, and the races that decide control — Texas, Ohio, Maine, Alaska, Kansas, Michigan and Minnesota — are close enough that turnout in the final three days, and the counting of mail ballots after Tuesday, could still change the story.

What Happens Next

Trump will hold restored rallies in Kansas and Iowa on Sunday, November 1, after a weekend of tele-rallies across roughly 19 states, while Democrats close with their own final appeals. Polls will close across the country on Tuesday, November 3, with Senate control likely to hinge on late-reporting states and, in Alaska's ranked-choice system, on how second preferences distribute.

Forecasters will update their models once more before Election Day, and any final polls in Texas, Ohio, Maine and Alaska will move the headline number again. If the polls hold, Democrats will enter election night as clear favourites; if they are off by a few points in the same direction, the night could look very different. Either way, Silver's update has changed the language of the race: from "too close to call" to a contest with a favourite — and a warning attached.

Frequently Asked Questions

What is Nate Silver's Senate forecast for 2026?

Silver's FLIPR model gives Democrats a 79 percent chance of winning control of the Senate in the November 3, 2026 midterms, up from 75 percent. A variant of the model reported by Newsweek put the figure at 77.7 percent.

How many Senate seats do Democrats need to gain?

Democrats need a net gain of four seats to move from 47 to 51 and take the majority. At 50-50, Republicans keep control because Vice President JD Vance casts the tie-breaking vote.

Which states will decide Senate control?

The key battlegrounds are Texas, Ohio, Maine, Alaska, Kansas, Michigan, Minnesota and North Carolina. Silver specifically warned that Democratic leads in Texas, Ohio, Alaska and Maine could flip on a normal polling error.

Does a 79 percent forecast mean Democrats are certain to win?

No. A 79 percent chance means Democrats win in roughly four out of five model simulations and lose in roughly one out of five. Polling error, turnout and late-deciding voters can all change the outcome.

What about the House of Representatives?

Silver said Democrats remain strong favourites to regain the House, without giving a revised probability in the comments cited. CNN's Harry Enten projected Democrats could win 223 House seats, above the 218 needed for a majority.

Related coverage: CNN Senate Polls: Democrats Lead in Michigan, Texas, Maine and Kansas | Trump Kansas and Iowa Rallies Back On After Cancellation | Trump Daytona Protesters Heckled Out as President Snaps