Prediction markets’ Wisconsin primary failure has rattled two of the industry’s biggest platforms, after Polymarket and Kalshi each gave progressive candidate Francesca Hong a 95% chance of winning a Democratic governor’s primary she went on to lose.
David Crowley, the Milwaukee County executive backed by outgoing Governor Tony Evers and other establishment figures, defeated Hong by a margin priced at roughly 0.4 to 0.6 percentage points in the 11 August contest.
Both platforms had aligned with traditional polling firms, which had also forecast an easy Hong victory. When votes were counted, Crowley prevailed by less than half a percentage point.
Polymarket deletes boastful post after prediction markets Wisconsin primary upset
The result prompted Polymarket to delete a post on X that had put Hong’s chances at 96%, even noting she lacked endorsements from prominent Democratic socialists including Bernie Sanders and Alexandria Ocasio-Cortez.
Kalshi’s founders pushed back on critics. Chief executive Tarek Mansour argued in a social media post that the outcome alone did not invalidate the forecast.
‘5% is not 0%,’ wrote Kalshi founder Luana Lopes Lara, pointing out that a 95% probability still gave Crowley roughly a one-in-20 chance of winning.
Crowley had faced earlier polling deficits but consolidated support from moderates and undecided voters in the final days of the campaign, according to the Polymarket margin-of-victory market.
A record under pressure
The Wisconsin result is not the first time the platforms have mispriced an election, though earlier misses came with wider uncertainty bands.
In June, both Kalshi and Polymarket gave reality-television personality Spencer Pratt about a 75% chance of advancing from Los Angeles’s nonpartisan mayoral primary to the general election. He finished third instead.
The month before, Kalshi gave Representative Thomas Massie a similar probability of winning the Republican primary in Kentucky’s 4th Congressional District. Massie lost to Trump-backed challenger Ed Gallrein.
Prediction markets had built their reputation in the final stretch of the 2024 presidential campaign, when Polymarket and Kalshi were widely credited with giving Donald Trump an edge in a race many conventional polls called a tossup. Their influence has since reached mainstream media: CNN named Kalshi its official prediction-markets partner, and Dow Jones plans to incorporate Polymarket data across The Wall Street Journal, Barron’s and MarketWatch.
Statistician and political forecaster Nate Silver cautioned against over-relying on the platforms. ‘I think prediction markets are cool,’ he wrote in a recent social media post. ‘But people should stop treating them as magic, and I don’t think they’re a good substitute for polls, or belong in models.’
The Wisconsin contest reinforced a point forecasters have long made: a 95% probability is a forecast, not a guarantee. Crowley’s win fell squarely within the 5% tail the markets themselves acknowledged.
