The slow decline of traditional polling has left politicos, who are always desperate for numbers, in a lurch.
Two venture-backed prediction market companies are offering nothing short of magic to meet those needs: real-time odds on who will win elections, and even projections of how much they will win by. As these companies lean into aggressively gamifying — or rather, gamblifying — U.S. politics, political operatives are warning campaigns not to take them seriously.
“The thing we have to remember is they’re not in any way representative of the voters that decide elections,” said Steve Schale, a Democratic strategist. “It is sort of like degenerate gambling for political hacks.”
Schale likened polls to “heroin” for people in politics. Prediction market projections offer more numbers for that demographic to obsess over, but that doesn’t make them helpful to campaigns, he argued.
Trending
Elections are big business for prediction markets. South Carolina’s Senate Republican primary runoff between Sen. Darline Graham and Rep. Ralph Norman saw more than $25.7 million in trading activity on Kalshi. The platform projected Graham’s odds at around 56% when polls closed, and she won.
Prediction markets, which claim to reflect the “wisdom of the crowd,” have missed the mark several times this primary season though: Last Tuesday, Kalshi’s market completely floundered in Florida’s Senate Democratic primary — it projected that insurgent democratic socialist state Rep. Angie Nixon had single-digit chances of winning. She won by a lot: The Associated Press called the race right as polls closed in the state.
In Wisconsin’s Democratic gubernatorial primary on Aug. 11, David Crowley pulled off an upset, even though Kalshi projected for weeks a single-digit chance of him winning.
The day before Abdul El-Sayed so narrowly clinched his victory in the Senate Democratic primary in Michigan, Polymarket projected his chances of winning at more than 98% — and the market on his victory margins predicted it would be a blowout, with him winning by at least 15%. That mirrored polls ahead of the election, which showed El-Sayed with a double-digit lead over his opponent, Rep. Haley Stevens. He won by less than 1 percentage point.
“It seems to me now that they’re much more reaction markets than prediction markets,” said Eddie Vale, a Democratic political consultant who works with Crowley’s campaign. “The theory was, if people had to put money or skin in the game, sort of the collective wisdom and knowledge was going to help predict things, but it’s pretty clear that people are just betting in reaction to seeing the news, seeing a new poll.”
“Anybody that has experience on campaigns knows that they’re useless,” Vale said of prediction market odds.
Prediction market platforms are forcing their way onto social media feeds through partnerships, including with newsrooms like CNN and Fox, and political influencers like Gunther Eagleman, a conservative poster with 1.7 million followers. Kalshi and Polymarket, the two biggest prediction platforms where people can trade event contracts, futures and derivatives on real-world events, are leaning into the political realm — both even have dedicated midterms betting hubs.
But the people who actually work around campaigns and politics have balked at some of the marketing.
Adam Carlson, a progressive pollster and prolific political poster, said he’s been approached with offers for paid partnerships with prediction markets. He said that while he personally finds prediction market odds “interesting,” he never checks on any candidate’s odds of his “own free will.”
“Polling can exist without prediction markets. Prediction markets cannot exist without polling,” Carlson said. “There is some value to them directionally, but there are some very clear weaknesses … It’s just more like a conversation starter if you’re not involved in the campaign. Or, it’s a discourse generator.”
One libertarian Republican strategist, Liz Mair, told NOTUS that she does see prediction markets changing politics because more people are being exposed to them — which can change how operatives think about running campaigns.
“The main change I’m seeing is that a lot of people who should know better are treating the odds offered by prediction markets as a substitute for polling. That’s a huge mistake,” Mair said in an email to NOTUS. “While there’s a lot wrong with polling … there’s zero reason why the opinions of a bunch of randos that skew prediction markets should be treated as remotely indicative of outcomes.”
Sure, polling has its issues, as both pollsters and political operatives conceded in interviews for this story. But they still generally have transparent methodologies, and strive for representative samplings that provide campaign operatives with nuanced, actionable data about where they’re falling short.
Only 14% of Americans report placing bets on prediction markets “sometimes” or “frequently,” according to a Navigator Research survey conducted in April. And those who do tend to be young men.
Prediction markets and polls also measure different things: Polls often ask people who they’d vote for, whereas markets ask who people think will win. For people running campaigns, the markets are more of a vibe check than a prognosis.
“A real poll is digging deep into their opinion on certain issues. That’s what drives things: ‘How do you feel about the economy? How do you feel about the price of gas? Should we be at war in Iran?’ Yada, yada, yada — a good pollster can read through all those and pull a result out. Well, none of that is in Kalshi,” Jacob Perry, a Democratic strategist, said.
Elisabeth Diana, Kalshi’s head of communications, said the platform is “more, actually, accurate than polling,” and pointed NOTUS to a Washington Post analysis of 2,000 primary election markets that found that despite several high-profile misses, markets were generally accurate, meaning candidates given a 75% chance of winning won around 75% of the time. The company has also done its own research to argue that its markets are accurate.
“We see prediction market forecasts as an effective complement to polling — it’s an additional data point that surfaces wisdom-of-the-crowd data, and it’s historically very well-calibrated,” Diana said in a statement.
The company has paid people to post about the betting platform on X, marketing it as something that could replace polls entirely, in an apparent effort to capitalize on people’s frustrations with traditional polling.
“Kalshi just released its new Midterms Hub, and it may be the final nail in the coffin for traditional polling,” Overton News, a news aggregator and Kalshi partner, posted on X. “The same pollsters who were embarrassingly WRONG about the 2024 election.”
While operatives and strategists warn against making campaign decisions based on the platforms’ projections, some said it can be used to fundraise, since it can make donors feel better about a candidate’s viability. The Democratic Senate candidate in Alaska, Mary Peltola, used her Kalshi odds to argue that her campaign has “a 57% chance to beat the DC establishment and take back Alaska” in a post on X earlier this month. The Senate Majority PAC, a super PAC aligned with Minority Leader Chuck Schumer, boasted in an email fundraising bid last month about Kalshi’s projections showing that Democrats have a chance of taking back the Senate.
“Because they update hourly, well, literally, minute by minute, they’re a very attractive, shiny object to rely on. But we don’t have decades of experience with how accurate they are, in the same way that we have many decades of experience with polling, both for good and bad,” said Charles Franklin, the director of Marquette Law School Poll. “It’s a wildly unrepresentative sample of the public.”
Given that prediction market bettors are largely on the younger side, could operatives’ dismissal of their worth speak to a generational gap? Mandy Zhang, a Gen Z Democratic digital strategist, said that while her friends use prediction markets for fun, nobody treats them like they’re prophetic.
“Digital teams do pay attention to where their candidates are sitting on these platforms, mostly as a quick pulse check or another data point. But I don’t think anyone I’ve worked with treats them as the ‘holy grail’ of political forecasting. They’re something people look at and talk about, but not something campaigns are necessarily building their strategy around,” Zhang said in a message to NOTUS. “Working across different campaigns this year, I’ve seen firsthand that they can be pretty wrong.”
Matthew Bartlett, a Republican strategist, said that if he were advising a campaign this year, he would tell them: “My short advice? Stay away.”
“You have so many people, just like sports betting is almost as big as sports, the betting markets are almost as big as politics themselves, and when you’re talking about real, human events such as sports or elections, they’re prone to disruption,” he said. “Win by the polls, die by the prediction markets.”
Author
Contributor
