Physical Address

304 North Cardinal St.
Dorchester Center, MA 02124

The Polymarket Bubble: Everyone Is Betting on the US Election

On Sunday, a quote-tweet from Musk sent former president Donald Trump’s election odds skyrocketing on the election betting platform Polymarket.
“Trump now leading Kamala by 3% in betting markets,” wrote Musk. “More accurate than polls, as actual money is on the line.”
The day after Musk’s tweet, Trump was up by nearly 10 points on the platform.
These aren’t actual polls—Polymarket is a decentralized betting market that operates similarly to sports betting apps like DraftKings, except bettors can make wagers on almost everything. This year, however, one of the most popular markets is the one predicting the outcome of the 2024 presidential election. The company is leaning into it, endlessly posting election updates on its X account and reply-guying Musk.
It’s not your average politics newsletter. Makena Kelly and the WIRED Politics team help you make sense of how the internet is shaping our political reality.
Polymarket has had a lot of success this year. The pro-Trump billionaire Peter Thiel and ethereum founder Vitalik Buterin announced they had raised around $70 million in funding for the company this past May. The site’s monthly traffic has grown to around 20 times more than what it was at the time of the investment, according to data from Similarweb.
Betting markets also secured a big win earlier this month when a US appeals court gave Kalshi, a predictions startup, the green light to offer markets on which party would take over Congress next year. While the case hasn’t been completely resolved, it likely won’t be decided before election day. This means election markets are here to stay, much to the joy of political influencers and elites like Musk online.
As of Wednesday afternoon, other betting sites PredictIt and Kalshi had vice president Kamala Harris up anywhere from one to three points. Unlike Polymarket, these alternatives limit bettors to placing no more than $1,000 on elections. They’re also subject to US regulations. Being located offshore, Polymarket is free from those same restraints, and bettors can place millions of dollars in wagers on the election outcome without anyone knowing who they are. Which means even if US-based bettors are formally banned from the site, all anyone needs to do is download a VPN to get around the rules.
Still, Polymarket has made itself known across the aisle. In August, Polymarket hosted a party at the Democratic National Convention. The guest list for the event was impressive: It included White House domestic policy director Neera Tanden, Olivia Rodrigo’s ex-boyfriend producer Adam Faze, and Senator Kirsten Gillibrand, who has supported a number of pro-crypto bills in the Senate.
This week, I spent some time reading through the Polymarket Discord server to get a sense of the folks most invested in the platform. There were people DMing each other what looked like entirely fabricated polls supposedly before they dropped publicly and suggesting countless other election hypotheticals to place their money on. One user wrote that Polymarket was “essentially sponsoring the US election” next to an emoji of Joe Biden giving a thumbs up.
Rajiv Sethi, an economics professor at Barnard College at Columbia University, says the jury is still out on whether prediction markets are more accurate than traditional polls. But even if they were totally reliable, Polymarket wouldn’t be the most accurate of all the markets available.
“In PredictIt, you get a relatively low volume, but no trader can dominate the market. It’s very hard to manipulate PredictIt, and it’s hard for any trader who may have beliefs that are out of step with the average to have a disproportionate impact on the price,” says Sethi. “That’s not what you see on Polymarket.”
With nearly three-quarters of all the traffic going to Polymarket’s website coming from men, the platform’s clearly not inclusive of all demographics either.
But maybe being correct isn’t even the point. Already, Polymarket odds are being used as evidence that Trump is beating Harris, and Musk and other Trump acolytes are using the odds to pump up their base. I’ve seen Democrats celebrating Harris’ odds of winning the popular vote on the site too. All of this has the potential to legitimize these results as viable evidence for conspiracy theorists questioning the outcome of the election.
“Right before the race gets called, Trump and his fans are going to say Polymarket knew the truth and they silenced it. It doesn’t matter if it’s right. It doesn’t matter if it makes any sense in those few hours after the election is called,” says Mike Rothschild, an author who writes about conspiracy theories. “People are going to be looking for any sort of evidence that there was a steal, that there was rigging, that there was the blue ballot dump at three in the morning. And if they can’t find it, they’re going to make it up.”
Like Scott Nover mentioned in Slate this week, if right-wing agents like Charlie Kirk will share random text messages as proof of a weak hurricane response from the Federal Emergency Management Agency, it’s not hard to imagine them leveraging Polymarket results in their favor as well.
“You can take anything that’s going on and turn it into evidence of the thing that you believe,” says Rothschild.
Hello! I’m Tim Marchman, WIRED’s director of politics, security, and science. And I wanted to take a moment to write about a new development in the world of John F. Kennedy assassination research. It’s surprisingly relevant to the presidential campaign—both directly and in the sense that the underlying pattern of government opacity at issue here does something to explain why this election is so defined by conspiracy theories about everything from microphone earrings to research of the ionosphere.
When he accepted Robert F. Kennedy Jr.’s endorsement in August at a rally in Arizona, former president Donald Trump announced that if elected, he would appoint a presidential commission to release all government documents related to the assassination of JFK. “This is a tribute in honor of Bobby,” he said.
Trump has made similar promises before. In 2017, he tweeted, “I will be allowing, as president, the long blocked and classified JFK FILES to be opened,” only to block the release of thousands of sealed and redacted records five days later, citing “potentially irreversible harm to our Nation’s security.” It appears, though, that there are previously unknown and highly intriguing records which he or Kamala Harris could order to be released, or at least investigated, if elected. The question now is whether they will.
This week, journalist and JFK assassination expert Jefferson Morley published a report in which, citing an anonymous whistleblower, he makes an extraordinary claim: that the CIA, at least at one point, maintained a sensitive compartmented information facility in a building near Dulles Airport in Virginia that held JFK records and that among them was a gray plastic video case labeled “Oswald in Mexico” or “Oswald in Mexico City” and dated September 1963.
This is significant because it would demonstrate that for more than 60 years, the CIA has been actively concealing what it knows about Oswald’s trip to Mexico City to visit the Soviet and Cuban embassies in fall 1963. The ambassador to Mexico at the time said that the CIA had Oswald under surveillance; the CIA’s station chief suggested in his memoir that this surveillance had yielded photos of Oswald; two retired CIA officers told congressional investigators that the chief had actually shown them photos; and Morley interviewed a former CIA employee who said that his mother, a CIA employee who had overseen surveillance of the Soviet embassy in Mexico City, had herself taken photos of Oswald.
Despite all this, the CIA has maintained, including to Congress, that it did not take and does not have photos of Oswald’s visit to the embassy. (Interestingly, it does not deny the accuracy of Morley’s reporting; the agency’s statement to the journalist, reading “CIA believes all of its information known to be directly related to the assassination of President John F. Kennedy in 1963 has already been released,” contains enough qualifiers to not say anything at all.)
This isn’t an academic matter; researchers have long held that information about Oswald’s visit to Mexico could resolve several mysteries related to the assassination and why the CIA has lied to Congress and the public. The most prosaic explanation is that the CIA, embarrassed that the president was assassinated by a man who had at one point defected to the Soviet Union and whom it was surveilling, has engaged in a campaign of ass-covering for more than half a century. More exotic theories include that the CIA concealed evidence of Soviet complicity in the assassination in the belief that exposure of this would inexorably lead to a nuclear holocaust.
Whatever the truth is, it can’t be known without the public being given access to government records, which is why in 1992 Congress passed a law mandating that these records be released in full by 2017, with a limited exception for records that not only pose an identifiable harm to national security but one that outweighs the public’s interest in full disclosure, as determined by the president.
How releasing documentation of Beatlemania-era intelligence operations could identifiably damage US national security interests is unclear, but despite his boasts, when he had the power to order the full release of these records, Trump continued to conceal them. President Joe Biden, meanwhile, last year published a memo asserting that the legally mandated review and release of all government records was complete—while noting that many records would remain secret because of the unspecified harms they could cause.
The Trump and Harris campaigns and the CIA did not respond to WIRED’s requests for comment.
The CIA seemingly lying and neither Republican nor Democratic administrations being particularly committed to transparency in this area have caused their own very real harms, though, as seen in the spread of conspiracy theories about how would-be Trump assassin Thomas Crooks was himself a CIA asset. While on their face absurd, such theories are in many ways simply a consequence of the public understandably lacking trust in a government that conceals the truth from it.
What do you think about all of this?
Send your thoughts to [email protected].
Want more? Subscribe now for unlimited access to WIRED.
🔗 Democrats dominate spending on political ads—except on Elon Musk’s X: Democrats are outspending Republicans on most digital advertising platforms, but the GOP dominates on Elon Musk’s X. Republicans have spent three times more on X ads than their Democratic opponents. (The Washington Post)
🔗 Inside the Patriot Wing: WIRED contributor Tess Owen has a new story out this week documenting her year of conversations with David Dempsey, a January 6er sentenced to 20 years in prison. Behind bars, he’s only become more ardent in his beliefs. (New York Magazine)
🔗 Porn Industry Jumps Into the Presidential Campaign, Targeting Project 2025: A group of porn actors are buying ads in swing states going after the Heritage Foundation and Project 2025 for targeting the industry. (The New York Times)
On the WIRED Politics Lab podcast this week, it’s the Elon Musk episode! Leah Feiger, Vittoria Elliot, and David Gilbert discuss the Butler rally with Trump and Musk, as well as Musk’s political evolution and what his involvement could mean ahead of November 5. You can listen to it here!
The podcast has also been nominated for a Signal Listener’s Choice Award! Please vote for us here.
My suggestion this week is for another newsletter! There’s less than a month left until Election Day and we’re going to need all the help we can get to make sense of the results. Kyle Tharp, the author of the FWIW newsletter, is the first person I call when I need insight into what the Democrats are doing online. I highly recommend subscribing to it–you can sign up for it here.
That’s it for today—thanks again for subscribing. You can get in touch with me via email, Instagram, X, and Signal at makenakelly.32.

en_USEnglish