People are betting on what’s going to win Game of the Year at The Game Awards. Well… technically, most of them aren’t betting. There are odds available at a couple of online bookies, Virgin Bet and BetMGM. (Both have Clair Obscur: Expedition 33 at 1/10 to win, and second-place runner Hollow Knight: Silksong at 5/1.) But this isn’t where the real action is.
Instead, wagering on Game of the Year has taken off on online prediction markets like Kalshi and Polymarket. These are, legally speaking, not gambling at all. They are peer-to-peer financial markets where users trade yes/no “contracts” on the outcome of real-world events. As such, there is no “house” setting odds that users are betting against. The market just takes a cut of the transactions, and the prices and payouts are set by the trading, as with a real financial market. If a lot of people are buying, the price goes up; if they’re selling, it goes down. They’re a bit like futures markets, only they’re about predicting events, not the price of goods.
You can buy a contract on Clair Obscur winning Game of the Year at 92 cents; if it wins, you’ll get back one dollar. If you prefer an outside bet, you can spend 2 cents on a contract for a Donkey Kong Bananza win, with the same one-dollar payout. More risk, more reward. It’s not gambling — but it also absolutely is. (Kalshi has made headlines because its status as a financial market essentially circumvents U.S. states’ bans on sports betting, although that status is now threatened by a recent judicial ruling that Kalshi is subject to gambling regulation in Nevada.)
At time of writing, with almost two weeks to go until The Game Awards, Kalshi has seen over $2 million of trading volume on its Game of the Year market. At Polymarket, Game of the Year trading has exploded since the nominations were announced. At the start of November, the market was recording less than $1 million of trading volume. Now, Polymarket is claiming almost $38 million of trading on Game of the Year.
These numbers sound high. In fact, they’re pretty average.
“That is a… decent number,” Jack Such, who does press relations for Kalshi, tells Polygon about Kalshi’s Game of the Year market in a video call. “It’s OK. But some markets have hundreds of millions.” Kalshi did more than $100 million of trades on the New York City mayoral election, and “something like” $550 million on the Presidential election.
Such points out that it’s not over yet, as trading typically intensifies as the deadline for the market approaches. “People like getting a resolution quicker,” he says. Kalshi’s Presidential election market was only open for three weeks, but half of its massive trading volume was in the 48 hours before the election. “There is this kind of exponential curve of trading volume up to when the event actually happens,” he says.
That said, the situation with Game of the Year in 2025, where there is an overwhelming favorite, can depress trading. This is a difference between a prediction market and betting; traders can always choose to sell their position before an event takes place, and potentially make a profit on market movements regardless of how the event resolves. But if the outcome is quite certain, traders are less likely to do that, which means less trading volume. “When there’s a market with an overwhelming favorite, we typically have people just buy and hold their position all the way through,” Such says. You could buy Clair Obscur contracts in April for 40 cents and sell them in October for 80 cents, or you could just wait until its seemingly inevitable win and pocket the full dollar. (Pity the traders who bought Grand Theft Auto 6 at 50 cents back in February.)
Unlike a sports book, the prices on prediction markets are driven by the trading, rather than by an expert who’s motivated to ensure the house wins. In a sense, they’re an example of the wisdom of crowds. Such claims prediction markets are “incredibly accurate.” A statistical analysis method called calibration measures the probability generated by the market against how often those outcomes are resolved; ideally, markets priced at 20 cents should resolve 20% of the time, 50 cents 50% of the time, 80 cents 80% of the time, and so on. A perfectly calibrated market should resolve as a straight-line graph of prediction against outcome. “When you look at Kalshi, it’s pretty much a straight line,” Such says. “It’s within a statistically insignificant difference of being the closest thing to a crystal ball that there is.”
That said, prediction markets are better suited at calling the outcome of some events than others. They’re good at predicting the outcome of political elections — perhaps better than traditional polling — because there’s a similarly large pool of people deciding the outcome in an open process. But Such gives the example of a papal conclave as the opposite: a black-box event decided by a tiny number of people, with no information going in or coming out during the process.
“Prediction markets are aggregators of information. So when there’s not a lot of information available, they’re not very accurate,” he says. The markets for the 2025 conclave that elected Robert Prevost as Pope Leo XIV “were not very accurately priced,” he says. “A lot of people who had [Prevost] made a lot of money. He was priced at 1 or 2%.”
In theory, cultural awards like The Game Awards, with their limited and specialized voting bodies, are more like papal conclaves than political elections. But Such theorizes that traders are drawn to them because of their innate love of culture and desire to engage in it and show off their expertise. “Unlike something like politics, where there are tangible events that happen all the time, like bills passing or endorsements or elections, or sports where there are, like, a hundred sports games every week, there are not that many measurable, tangible events in culture,” he says. “Awards are kind of the closest proxy for that. So for everyone who loves pop culture, whether it be movies with the Oscars or music with the Grammys or video games, awards are the best thing for them to engage in the markets on.” Kalshi has other video game markets, around such things as Metacritic scores or release dates, but “historically awards have just been what people gravitate towards.”
Unlike a bookie, Kalshi doesn’t take on much risk with its markets. It just takes a cut of the trading, regardless of the outcome. But it still needs to invest in expertise to create the markets; in the case of Game of the Year, ensuring that contracts on all the right games are available to trade, for example. This work is done by a dedicated market team. “We have to be involved, because every market you see on Kalshi requires a pretty extensive legal contract,” Such says. But Kalshi does monitor feedback and take suggestions from users, if enough people say an obvious option is missing.
Such says Game of the Year is one of the oldest markets on Kalshi; he’s not sure of its genesis. But in general, he says gaming markets are among the most highly requested by the trading community. “It’s very clear from our community that this was a space we were under-invested in in the past, we got a lot of consumer demand to put [more gaming markets] up.”
Perhaps it’s unsurprising that gamers are drawn to an online space where they can use their expertise to play games with risk and reward — or, to flip it, that predictions traders are drawn to gaming. The communities seem like they’d have a natural overlap. But really, predictions markets are just answering a deep instinct to bet you know better, and the Game of the Year race is as susceptible to that instinct as anything else. “If you think you’re right, you think you’re smarter than [the] market, you want to make money,” Such says. “And we let people do that.”
