Photo-Illustration: Curbed; Photos: Getty

The prediction-market platform Polymarket recently asked users to guess where New York City’s median home value would fall on March 1. The rules were as follows: “This market will resolve according to the median home value for all property types in New York City, New York on March 1, 2026. If the reported value falls exactly between two brackets, then this market will resolve to the higher range bracket.” Users with names like Outgoing-Canoe and BR43 silently placed their bets. (When it comes to chatty gamblers, most of the daily action seems to be happening around the city’s weather. “Check the underground weather website, it just updated,” a user called Ujusylviia advised on a bet for March 9. “The highest temperature on 7th was 47 so good luck.”) The introduction of a monthly market around the median value of depressing one-bedrooms in the Financial District and fantasy brownstones in Brooklyn Heights was part of a new set announced in mid-January allowing the various “sharps” of the site to try and predict home prices in “high-liquidity cities” like ours. So, of course, they did.

And why not? Polymarket is in the business of transforming the news cycle into a series of hedgable data points; real estate, with all its possible variables and idiosyncrasies, is a natural extension. Housing prices are basically a complex math problem themselves. In the manic universe of Polymarket, any of the site’s 1.7 million trading addresses can, within the space of a few seconds, bet on the specific words uttered in a forthcoming podcast, if the U.S. will soon confirm aliens exist, and whether Palantir’s move to Miami will push home values in the metro area significantly higher. (Until quite recently, one could also wager on the chances of a global nuclear war.)

Some of the people (and entities) most likely to profit from the new market focused on New York real estate are investors looking to scoop up or sell properties at the correct moment. The others are the Polymarket obsessives with their endless cycles of hedges and shorts. (Given Polymarket’s stateside regulatory issues, many of the accounts seeking to capitalize on New York’s housing market likely originate overseas. Some  are bots.) Most of the betters using the platform lose money, though a few actual people do make fantastic amounts of it turning every conceivable future event into a wager.

And more than making money, they’re allegedly divining the future. According to the platform’s boosters, Polymarket’s capacity for prediction — an occult, crowdsourced mixture of insider information, public sentiment, and broad cryptocurrency-backed vibes — can outstrip more conventional analysis: Prediction markets “represent a paradigm shift in how views are expressed and truth is identified,” said the CEO of the housing data startup Parcl after the real-estate initiative’s launch.

This is apparently true from the perspective of some buyers: According to Heather Taylor, who heads up a commercially focused real-estate firm in Colorado, some entry-level investors are watching Polymarket’s predictions to decide how to proceed. “What they felt was that if people were making these predictions, but also willing to put money on the line, then it was a really solid indicator as to what they felt was going to happen next,” she told me. “I guess if you have a cohort of people that are betting real money, then you kind of feel more safe about it,” whether those dollars are flowing to gamble on home prices or the chances Trump will send ground troops to Iran.

Within the hermetic subculture of the Polymarket fluent, burgeoning crypto influencers and Youtube investment accounts praised the new market in ChatGPT-inflected triplets: One user wrote on X that “a market worth tens of trillions is suddenly something you can express a view on in a few clicks,” predicting that “big, slow, real-world markets are becoming tradable options.” In one post, someone wrote that real-estate prediction markets “feels like hedging IRL without touching my mortgage,” describing the new markets as “pure crowd signal cutting through Zillow noise.”

“No leverage, no decay, no property taxes,” they continued. “Just clean yes/no on whether prices climb by quarter end. If you’re sitting on equity but hate the illiquidity, this is the synthetic play we’ve needed.” By which they meant, I guess, that prediction markets can extract smart money out of nothing — even if that nothing actually is the relative value of your apartment in midtown.

For a more measured take (spoken in human language) I talked to Dustin Gouker, an analyst and writer who publishes newsletters about sports betting and prediction markets. He said he was far from evangelical about Polymarket but thought the real-estate data culled from thousands of people with real stakes — contractors, homeowners, retailers — could be “useful to the world.”

“Other than owning real estate, there’s not a great way to project how things will go — or to economically hedge on what the market is going to do,” Gouker says. For instance, a person could “worry that the price of my house isn’t going to go up, so I can use a prediction market to trade on that outcome.” He likened Polymarket’s real-estate bets to election-focused prediction markets in which individual polls were often less accurate than Polymarket’s synthesized probabilities over time.

Given the platform’s anonymous, blockchain-based interface and the general volatility of its users, it’s impossible to say what influenced the February wagers on how much it would cost to buy a home in New York come March. But when Zohran Mamdani proposed a property-tax hike, predictions shuffled wildly. And around the time Polymarket posted its own dubious breaking-news update that “searches for ‘can’t sell house’ reach historic highs,” odds the median home value would drop below $575,000 briefly, if dramatically, spiked.

In the end, the winnings on the March 1 bet were relatively meager. When the prediction market closed, the exact number of a median home price at precisely 11:59 p.m. on February 28 (as determined by Parcl) was $590,110.00. A user named cry.eth2 placed the most money on Polymarket’s pre-determined bracket ( $590,000 to $595,000) and won a hundred bucks.

A number of Polymarket’s power users believe cry.eth2 is a bot, the kind an increasing number of Polymarket users deploy to analyze variables and quickly automate trades: Documentation elsewhere describes cry.eth2 as a simulacrum of a “high-conviction, venture-style trader, specializing in acquiring extremely cheap ‘yes’ shares across a vast range of future events.” The account made over $75,000 in the last month. According to Polymarket’s internal tracking mechanism, since it was deployed it has netted $222,898.14 for whoever is pulling the strings — including by correctly predicting the election of Lee Jae-myung as the president of South Korea and Oklahoma’s NBA Finals win. But in case you’re worried about missing your chance, there’s always another opportunity to show up the machines. The Polymarket odds on April 1 housing prices are currently live.

Related