How Accurate Was Polymarket During the US Election?

Key takeaways:

Polymarket accurately predicted major outcomes in the U.S. election, outperforming traditional polls in key swing states. While Polymarket proved effective, it missed predicting the popular vote winner. Post-election, Polymarket’s activity dropped significantly, as attention shifted away from high-stakes markets.

In the run-up to the US election, there was quite a bit of cynicism toward Polymarket — a crypto-powered site where users can place bets on the outcome of future events.

The platform had loftily been described as a superior alternative to opinion polls, and a more accurate barometer of political reality than the mainstream media.

But critics had argued that its data had been distorted by wash trading, with a small number of whales placing outsized bets that skewed odds in Donald Trump’s favor.

With the world digesting news that Trump will serve a second term as president, Polymarket has not just survived its first major test, but passed with flying colors.

Opinion pollsters are now beginning a grim post-mortem to assess how they could be so wrong about the US political landscape. Right up until election day, survey after survey suggested that “basically a coin toss” separated Trump and rival Kamala Harris.

But at 9am ET on Tuesday, as Americans were beginning to cast their ballots, Polymarket’s live odds were already showing a distinct edge for the Republican candidate.

Polymarket odds on Tuesday morning

Polymarket’s edge continued as results were counted and verified.

The Associated Press news agency has had a role in declaring the winners of US elections since the mid-1800s, making calls when it’s mathematically certain there’s a victor.

And although AP’s results are used as verification before payouts are made, Polymarket argues its users knew the outcome in key swing states hours earlier.

In that graphic above, Polymarket says its equivalent of calling a race is when a candidate has more than a 95% chance of winning.

The site says it “called” North Carolina 90 minutes before AP — and the overall winner a full nine hours earlier. And while this wire service maintains that Arizona remains too close to call, it’s already a foregone conclusion in Polymarket’s eyes.

Swing state odds at 9 am ET on Tuesday

Overall, there was widespread expectation that the eventual victor in this hard-fought campaign would boil down to seven battleground states: Michigan, Wisconsin, Pennsylvania, Nevada, Arizona, North Carolina and Georgia.

No opinion poll or political pundit would have predicted that there would be a red sweep across them all — but as things stand, Trump has won five of them, and is ahead in the other two.

As the screenshot above shows, Polymarket’s odds accurately gave Trump an edge on Tuesday morning — putting his chances at 70% in Georgia, and 61% in the crucial state of Pennsylvania, which carries 19 electoral votes.

That being said, the site didn’t get every call right. When Cryptonews took these screenshots at 9am ET on election day, Harris was tipped to secure Wisconsin and Michigan. In the end, the 25 Electoral College votes from both states went to Trump, and each victory was fairly narrow.

Swing state odds at 9am ET on Tuesday

And powerfully illustrating how the wisdom of the masses isn’t always a crystal ball into the future, Polymarket had one big miss: which candidate would win the popular vote.

As I’m sure you’ll know by now, the popular vote is a different metric to the Electoral College — and refers to which candidate secured the most ballots nationwide.

It’s possible to get to the magic number of 270 electoral votes and win the presidency without having the support of most Americans, as happened with Donald Trump in 2016.

Interestingly, while Polymarket said Trump had a 61.7% chance of securing a second term as of Tuesday morning, just 26% believed that he would also win the popular vote.

The latest figures from NBC News suggest he in fact managed both — with the president-elect gaining 72.6 million votes compared to his rival’s 67.9 million.

That goes to show that there’s always an element of surprise that’s difficult to predict.

Source: NBC News

What Next for Polymarket?

After more than $3.5 billion was spent wagering on the US election, it’s fair to say that Polymarket looks a little bit more subdued now.

Attention has now turned to what Trump’s priorities may be after he’s inaugurated on January 20, but each of these markets have dramatically smaller trading volumes by comparison.

‘Top’ markets on Polymarket’s homepage now

Just $431,000 has been wagered on whether Trump will give Elon Musk a seat in the cabinet, with $155,700 bet on who the president-elect will pardon when he returns to power. Both are listed as “top” markets on its homepage.

This chart shows how November 7 is already shaping up to be a tiny fraction of the frenzied activity seen as the election reached its dizzying climax.

Source: Dune Analytics/@fergmolina

While the presidential election winner was by far and away the most popular market on Wednesday, Thursday’s focused on who might win the Super Bowl in 2025.

Despite Polymarket’s goal of differentiating itself from 24-hour news channels, it turns out there’s one thing they share in common: both only get attention when something’s happening.

And although this betting platform might have an easier ride from a regulatory standpoint during the Trump administration, it could face headaches in other jurisdictions.

Reports are suggesting that France could move to ban Polymarket altogether and block its domain, amid concerns it falls afoul of the country’s gambling legislation.

The post How Accurate Was Polymarket During the US Election? appeared first on Cryptonews.