The 2025 New York City mayoral race became one of the most heavily traded political markets ever built on Polymarket and Kalshi, with tens of millions of dollars wagered on who would run City Hall. The story that these markets told did not stay still. It swung from a Cuomo coronation to a Mamdani landslide, and along the way it exposed both the strength and the limits of using betting odds to forecast an election. This page walks through how the odds moved, what actually happened on the ballot, how the markets compared to the polls, and how you can read and trade election markets on Kalshi and Polymarket yourself.
Don't miss new tech stories on Google
Add Tech Insider once in the Google app and our stories appear in your news suggestions.
The short version: what happened in the 2025 NYC mayoral race
Zohran Mamdani, a 34 year old state assemblymember and democratic socialist, won the November 4, 2025 general election. He took 1,114,184 votes, or 50.78 percent, and became the first candidate to cross one million votes in a New York City mayoral race since John Lindsay in 1969. Former governor Andrew Cuomo, running on an independent line called Fight and Deliver after losing the Democratic primary, finished second with 906,614 votes (41.32 percent). Republican Curtis Sliwa came in third with 153,749 votes (7.01 percent). Turnout reached 2,218,647 votes (43.47 percent), the highest since 1969.
That outcome looks tidy in hindsight. It was anything but predictable when the cycle began. The prediction markets spent most of the spring pricing Cuomo as the near certain winner of the Democratic primary, then reversed hard in the space of a few weeks. Understanding that reversal is the whole point of studying these markets.
Spring 2025: Polymarket priced a Cuomo comeback
For much of the early campaign, Cuomo was the heavy favorite. He carried enormous name recognition, a large fundraising advantage, and a base of older and more moderate Democrats. Polymarket reflected that. At the end of May 2025, contracts on Cuomo winning the Democratic nomination traded around 93 percent, while Mamdani sat near 6 percent. Kalshi, which lists regulated event contracts in the United States, told a similar story. If you had glanced at either market in May, you would have concluded the primary was effectively decided.
The mispricing, if you want to call it that, was not obvious at the time. Cuomo led most public polls. The markets were doing exactly what markets do, which is to convert the visible evidence into a price. The problem is that the visible evidence was about to change faster than the polling averages could capture.
June 2025: the endorsement, the surge, and a market caught flat footed
The turning point came in early June. A high profile endorsement of Mamdani from Representative Alexandria Ocasio-Cortez, combined with a ground game that was mobilizing young and first time voters, shifted the race. Mamdani’s Polymarket odds climbed quickly. By June 12, some Polymarket contracts had Mamdani at roughly 42 percent and Cuomo at 57 percent, a stunning move from the 93 to 6 split just two weeks earlier. Mid June readings bounced around, with Cuomo still favored in the low 70s on some days as traders tried to price a genuinely uncertain contest.
On the eve of the June 24 primary, the two platforms had narrowed to something close to a coin flip. Polymarket at one point favored Cuomo by roughly 53 to 47, and both markets still listed Cuomo as the nominal favorite going into election night. Then the votes came in.
Primary night: the upset that the markets did not fully see
Mamdani won the Democratic primary in an upset. On the first round of the ranked choice count, he led Cuomo by roughly 43.5 percent to 36.4 percent, a margin large enough that Cuomo conceded that same night. When the ranked choice rounds were fully tabulated and confirmed on July 1, Mamdani had beaten Cuomo 56.39 percent to 43.61 percent, a gap of about 12 points.
This is the moment that matters most for anyone evaluating prediction markets. The markets had spent months with Cuomo as a lock, then narrowed to a toss up only in the final days, and still had Cuomo slightly ahead going into the count. The eventual 12 point Mamdani win was not what the closing prices implied. The markets moved because of the polls and the late momentum, not ahead of them. That is the opposite of the pitch that prediction markets are usually sold on, which is that they aggregate private information and lead the polls rather than follow.
Once the result was in, Polymarket contracts on Mamdani winning the nomination spiked toward 99 percent almost immediately. Prices reprice instantly on new information, which is a feature, but the sharp jump also underlined how much of the earlier confidence in Cuomo had been misplaced.
Summer and fall 2025: the general election market
After the primary, the contest reshaped. Cuomo relaunched as an independent on the Fight and Deliver line on July 14. Incumbent mayor Eric Adams, who had already left the Democratic primary to run as an independent back in April, formally withdrew from the general election on September 28, though his name remained printed on the ballot. Sliwa held the Republican line. The general election market on both platforms now tracked a very different question: could anyone stop Mamdani in a three way race.
The markets said no, and this time they were more aligned with reality. Through the summer and fall, Mamdani stayed the clear favorite. Polymarket had him around 82 percent in August. In mid September, analyst Nate Silver publicly argued Mamdani was actually a good buy relative to that price, suggesting the market was slightly underrating him. On Kalshi, Mamdani sat near 84 percent after Adams dropped out in late September. As election day approached, Polymarket contracts on Mamdani firmed into the low to mid 90s, reaching roughly 91 to 94 percent by the start of November. When the votes were counted on November 4, the contracts resolved to 100 percent for Mamdani.
The monthly rhythm of the Polymarket general election market is worth tracing, because it shows how a market behaves once the big question is essentially settled. In July, immediately after the primary, Mamdani traded around 95 percent while Cuomo sat near 3 percent. August brought the only real wobble, with Mamdani easing to about 82 percent and Cuomo ticking up toward 10 percent as the independent campaign found some early footing and a fresh round of polling tightened the picture. From September onward the market drifted back toward Mamdani, holding roughly 91 to 92 percent through October, with Cuomo stuck in the 6 to 7 percent range. Daily contract volume climbed across the same stretch, from the mid teens of thousands in July to more than 27,000 by the start of November, a sign that money kept flowing in even as the price barely moved. That combination, high and rising volume around a stable favorite, is typical of a race the market considers close to decided.
One recurring driver of the day to day noise was the crypto angle. Cuomo made public overtures to cryptocurrency voters and donors during the general election, and each time that news broke, traders on the crypto native Polymarket reacted, nudging Mamdani’s contracts up or down by a few points. None of it changed the underlying trajectory. The market treated these developments as marginal, and it was right to.
The general election market did what supporters of prediction markets want them to do. It identified the favorite early, held that view against noise, and closed very close to the eventual result. The primary market did the opposite. The same event produced both a success story and a cautionary tale, depending on which stage you look at.
Peak market-implied probabilities versus the actual outcome
The table below summarizes the highest implied probability each major candidate reached on Polymarket during the cycle, the specific market it applied to, and how it resolved. Peak figures are approximate and reflect intraday highs reported across the campaign rather than any single official close.
| Candidate | Market | Peak market-implied probability | Actual outcome |
|---|---|---|---|
| Andrew Cuomo | Democratic primary nominee | ~93% (late May 2025) | Lost primary; conceded on June 24 |
| Zohran Mamdani | Democratic primary nominee | ~99% (after June 24 result) | Won primary, 56.39% in final ranked-choice round |
| Zohran Mamdani | General election winner | ~100% (resolved Nov 4) | Won, 50.78% (1,114,184 votes) |
| Andrew Cuomo | General election winner (independent) | ~10% (August 2025) | Lost, 41.32% (906,614 votes) |
| Curtis Sliwa | General election winner | Low single digits throughout | Lost, 7.01% (153,749 votes) |
How the markets compared to the polls
The honest scorecard is mixed, and it splits cleanly along the primary versus general line.
In the primary, the polls arguably did the harder work. A final Emerson College survey showed Cuomo ahead on first choice support, roughly 35 percent to 32 percent, but its ranked choice simulation had Mamdani winning the final round, around 52 to 48. That simulation pointed at the eventual result more clearly than the closing market prices did. The markets, by contrast, spent months anchored to Cuomo and only converged toward a toss up at the very end. They followed the late polling and the momentum rather than anticipating the upset. For a mechanism that claims to beat polls, being dragged along by them is not a strong showing.
In the general election, the picture flipped. The polls and the markets agreed that Mamdani was the strong favorite, and both were right. Here the markets added value by staying steady through the campaign’s daily noise and closing very near the true outcome. Silver’s mid September call, that Mamdani was underpriced at 82 percent, also aged well.
The broader lesson is one that experienced traders already accept. Prediction markets are excellent at pricing races where the outcome is reasonably stable and the information is widely shared, such as a frontrunner cruising through a general election. They are much weaker at pricing volatile, low turnout, ranked choice contests where a late surge can rewrite the result faster than any signal can propagate. Treat them as one input alongside polls, fundamentals, and turnout modeling, not as an oracle.
How to bet on and trade election markets on Kalshi and Polymarket
The mechanics of these platforms are simpler than the jargon suggests. Below is a practical walkthrough that stays useful well beyond any single race.
Understand what a contract is
An election contract is a claim that pays out a fixed amount, usually one dollar or one unit, if a specified outcome happens, and zero if it does not. The price you pay is the market’s implied probability. A contract trading at 60 cents implies a 60 percent chance. If you buy at 60 cents and the outcome occurs, you receive one dollar, a gain of 40 cents per contract. If it does not occur, you lose your 60 cents. This is why the price and the probability are the same number expressed two ways.
Know the difference between Kalshi and Polymarket
Kalshi is a United States based exchange regulated by the Commodity Futures Trading Commission. It settles in dollars, requires standard identity verification, and lists a curated set of event contracts, including political and election markets. Because it is regulated domestically, it is the more straightforward on ramp for many United States residents.
Polymarket is a larger, crypto native platform that settles trades in a stablecoin. It typically offers deeper liquidity and a wider menu of markets, which is why its NYC mayoral market attracted such large volume. Access and availability depend on your jurisdiction and on the platform’s current rules, so confirm eligibility before funding an account. The two platforms often list the same race, and their prices usually track each other closely, though small gaps can appear.
Fund an account and place a trade
On Kalshi, you create an account, verify your identity, and deposit dollars by bank transfer or card. You then open a market, choose the outcome you want, and buy contracts at the listed price using a market order for speed or a limit order to set your own price. On Polymarket, you connect a compatible wallet, fund it with the supported stablecoin, and buy shares of the outcome you expect. In both cases your position value moves in real time as the price moves, and you can sell before the event resolves to lock in a gain or cut a loss.
Read the order book and the liquidity
Price is only half the picture. Check the order book, which shows how many contracts are available to buy or sell at each price. Thin markets can move sharply on a single large order, and you may pay more than the quoted price to fill a big position, an effect called slippage. The NYC mayoral market was liquid enough to absorb large trades, but many smaller political markets are not. Size your orders to the available depth.
Volume and open interest also tell you how much to trust a price. A contract trading at 90 percent on a market with a few thousand dollars of total volume is a much weaker signal than the same price on a market that has seen tens of millions change hands, because a lightly traded market can be moved by one determined participant. When you compare Kalshi and Polymarket on the same race, treat a persistent gap between the two prices as information. It usually points to a difference in who is trading on each platform or a temporary imbalance that may correct, and occasionally it flags a real disagreement worth investigating before you commit money.
Manage risk like a trader, not a fan
Three habits separate disciplined trading from gambling. First, decide your position size in advance and never stake money you cannot lose, because even a 95 percent favorite loses one time in twenty. Second, watch for the same trap the primary market fell into, where a comfortable price lulls you into ignoring late movement; set alerts and re-check the race after major news, debates, endorsements, and fresh polls. Third, remember that you can exit early. If new information turns against your position, selling at a loss is often smarter than holding to resolution and hoping.
Mind the rules and the fine print
Regulations around election trading change, and availability varies by state and country. Read each platform’s terms, confirm that political markets are offered in your location, and understand the fee structure, which can include trading fees or spreads that eat into thin margins. None of this is financial advice. Treat prediction markets as a tool for understanding probability and, if you choose, for taking calculated positions with money you have set aside for that purpose.
What the 2025 NYC race teaches about prediction markets
The single most instructive takeaway is that the same market can be both wrong and right in one election. Polymarket and Kalshi badly misread the Democratic primary, holding Cuomo as a near lock until momentum and polling forced a last minute repricing, and even then they did not fully capture the size of Mamdani’s win. Weeks later, the same platforms nailed the general election, keeping Mamdani as the clear favorite from summer through his 50.78 percent victory. Markets are a powerful aggregator of information, but they are only as good as the information flowing into them, and in a fast moving, ranked choice primary that information arrived too late. Use the odds, respect the odds, and never mistake a confident price for a certain outcome.
Frequently Asked Questions
Who won the 2025 New York City mayoral race?
Zohran Mamdani won the 2025 New York City mayoral election. His primary victory was the bigger surprise to prediction markets, which had priced other candidates as more likely earlier in the cycle.
Did prediction markets predict the NYC mayor result?
Markets moved toward Mamdani as real information arrived, and by election day they favored the eventual winner. Earlier in the race they underpriced his chances, which shows markets update rather than forecast perfectly.
Can you bet on mayoral elections on Kalshi or Polymarket?
You can trade many election outcomes on both platforms where they are listed. Whether a specific local race is available depends on the platform and your location, so check that the market is open to you.
Were the markets more accurate than the polls?
By the finish the strongest polls and the markets landed close together, but the markets adjusted faster as new information arrived. Neither tool is a crystal ball, and both showed wide ranges earlier in the cycle.
How do election prediction markets actually work?
Each candidate has a yes or no contract that settles at one dollar if the outcome happens and zero if it does not. The live price, between 0 and 100 cents, reflects the market implied probability of that outcome.
Related guides on Tech Insider
- Election betting odds explained
- Is Polymarket legal in the USA?
- Kalshi vs Polymarket
- What is Kalshi?
- Prediction markets vs sportsbooks
Sources and further reading
