As this year draws to a close, the rapid growth of prediction markets has attracted increasing attention from both crypto enthusiasts and those outside the circle. When discussing prediction markets with different friends, I often explain the difference between them and traditional gambling through 'differences in betting models' and 'the financialization of information.' As these conversations multiplied, I gradually realized that these two points are still not intuitive enough for friends who haven't been exposed to prediction markets. They even raised some very interesting and sharp questions, such as:

I can understand the difference between 'betting against the house' and 'P2P gambling among players,' but in essence, both models still involve real-money betting. Why does the U.S. government treat the two so differently?

If the lenient attitude towards prediction markets is because 'the financialization of information' has many benefits, then why hasn't this good thing become popular before? Why did it only explode in 2025?

There has been much discussion about the advantages and future prospects of prediction markets. Therefore, this article will explore prediction markets from more interesting angles.

Why Did Prediction Markets Only Explode in 2025?

Report data shows that the prediction market is expected to achieve an astonishing 400% growth rate in 2025, with total trading volume projected to grow from approximately $900 million in 2024 to $40 billion. The user base is also expected to increase 3-4 times, from about 4 million in 2024 to 15 million in 2025.

Retail players familiar with prediction markets should be well-acquainted with the hot topics over the past two years, such as the 2024 U.S. presidential election and the 2025 League of Legends World Championship. One might instinctively think that these hot topics drove the development of prediction markets.

But this is clearly not the only or core factor. Traditional gambling platforms also offer the same markets, and compared to the binary 'Yes/No' of prediction markets, traditional gambling markets have even more variety, such as 'point spreads.' Moreover, before the rise of Polymarket and Kalshi, there were precedents for prediction markets. The earliest available prediction market platform was the Iowa Electronic Markets (IEM), launched by the University of Iowa in 1988 for predicting the 1988 U.S. presidential election. Even prediction market platforms incorporating blockchain technology were not pioneered by Polymarket; Augur on Ethereum came in 2018.

Significant progress in regulation is an important factor, and the relaxation of regulations has contributed to the explosion of prediction markets on multiple levels.

First, the rollout of distribution channels. On November 25, the CFTC approved Polymarket's Amended Order of Designation, allowing it to re-enter the U.S. market as a Designated Contract Market (DCM). Kalshi also integrated directly into Robinhood and Coinbase this year. More importantly, after compliance, prediction markets have broader distribution channels in the U.S. than traditional gambling, covering all 50 states, while traditional gambling distribution channels only cover over 30 authorized states.

The more profound significance is that compliance clarifies the positioning of prediction markets as legitimate commodity derivatives rather than gambling, affirming the positive significance of prediction markets. This expands the distribution audience beyond traditional gambling enthusiasts to include investors and decision-makers. For ordinary people, seeing traditional media reports citing prediction market data or seeing internet search engines like Google directly indexing prediction market data creates a positive perception of 'legitimacy' that the cryptocurrency industry has sought for many years.

Second, policy friendliness has given institutional investors sufficient confidence, leading to rapid growth in the financing path for prediction markets. Both Polymarket and Kalshi completed a total of three new rounds of financing in 2025, with new financing amounts each exceeding $1 billion. This has given them better conditions to provide better products and liquidity.

Finally, the variety of event types in prediction markets has also been enriched. In 2024, Kalshi won a lawsuit against the CFTC, allowing it to list more types of event predictions, such as cryptocurrency-related events, and since January this year, it has expanded to sports events. The CFTC abandoned its appeal in May this year. Currently, sports event predictions account for about 90% of Kalshi's trading volume. A report by Eilers & Krejcik believes that, in the long run, sports event predictions will account for 44% of the total trading volume in prediction markets.

Of course, whether it's Kalshi, which has always strictly followed a compliant, off-chain path, or Polymarket, which took an offshore-then-compliant, on-chain path, the improvement of the prediction market products themselves, along with advancements in AI technology that have provided more complete related tools for the entire prediction market ecosystem, are all reasons why 2025 became the year of prediction markets. As mentioned earlier, the earliest on-chain prediction market, Augur, was previously criticized for its poor user experience and, after four years of silence, announced in March this year that it would be restarted by the Lituus Foundation. The popularity of anything requires waiting for the 'right time, place, and people.' The wheel of fortune turns, and prediction markets have finally borne fruit in 2025.

Why Does the U.S. Government Treat 'Gambling' Differently?

This is a very interesting question. Even I was initially confused—although prediction markets themselves do not act as the house, do not provide odds, and do not bet against users, we cannot say that peer-to-peer betting is not gambling. Otherwise, we could also say that playing poker is not gambling because the organizer does not play but only takes a rake, and the betting only occurs among the players.

At the same time, traditional gambling platforms, for their own benefit, have professional teams to analyze various event bets and try to provide reasonable odds. In the past, traditional media and institutional reports have cited odds from traditional gambling platforms for event reporting and analysis. Although prediction markets, compared to traditional gambling, can better reflect 'collective cognition' without the bias of a single team, and unlike traditional gambling platforms, they do not deliberately align odds with the public to increase overall betting volume and use the rake to offset losses caused by unreasonable odds settings, and indeed show higher real-time responsiveness and accuracy in some events, this does not seem to constitute an overwhelming advantage over traditional gambling platforms.

And whether for traditional gambling enthusiasts or ordinary retail investors in prediction markets, our grasp and analysis of events are, in most cases, certainly not as good as professionals, which means losses are common.

So, isn't the U.S. government worried about the harm caused by the gambling nature of prediction markets to the public? Of course it is, otherwise the compliance path for prediction markets would not have been so difficult for many years.

Internationally, some prediction markets operate directly under gambling licenses, such as the political prediction market on the UK's Betfair. Recently, U.S. District Court Judge Andrew Gordon in Nevada revoked an injunction protecting prediction market company Kalshi from state regulation and ruled that Kalshi's sports event predictions do not constitute transactions under the 'Commodity Exchange Act.' The judge believed that these prediction events are very similar to sports betting wagers and therefore fall under the regulatory purview of the Nevada Gaming Control Board and the Nevada Gaming Commission.

But the government's starting point is 'overall positive externality.' Just as the stock market allows the vast majority of retail investors to 'pay tuition' while greatly promoting overall social and economic operation, prediction markets, although essentially still gambling, also have key social benefits such as improving information efficiency and aiding decision-making. As long as the harms can be avoided as much as possible through sound regulation, we should not give up eating for fear of choking.

Of course, after reading this, you might still say that 'overall positive externality' is a pale rebuttal, and traditional gambling can still play a similar role. However, at least one point is undeniable: the mechanism of prediction markets themselves determines that the platform can only 'choose topics' and cannot 'rig the game.' Whether Polymarket or Kalshi, they can at most choose topics that people like to encourage betting and cannot manipulate odds.

Moreover, besides the U.S., governments in other countries and regions currently agree with your thinking. The U.S. is the only country in the world that distinguishes the positioning of prediction markets from traditional gambling and issues 'event contract' licenses for prediction markets. As the scale of prediction markets grows, controversies surrounding gambling will continue, and whether the U.S. government's attitude will change remains unknown.

Maybe, let the prediction markets predict it...?

A Paradise for Insider Traders?

This is a very interesting question. If asked whether insider trading in the stock market is good or bad, we would all say it's bad. But if the subject of this question is changed to prediction markets?

My colleague believes this is the value of prediction markets. The existence of insider information certainly creates unfairness in betting, but it also most accurately discloses information. Therefore, a true prediction market should be like Polymarket, built on-chain, anonymous without KYC.

Such views are actually not uncommon. For example, this tweet by @shafu0x received considerable recognition:

Players who hold this view believe that as long as the information leaked through insider trading has 'overall positive externality,' it is justified. They also believe that prediction markets cannot do without these insiders because without them, the accuracy of event predictions in prediction markets would drop significantly, leaving only those without accurate information to bet against each other.

This view has some merit, but I think prediction market platforms themselves would resist such positioning because, in the long run, it would harm retail trust. The entire market would essentially change from a platform encouraging individual research and insight, aggregating collective wisdom, into a one-sided 'slaughter' by insiders, which is not conducive to the development of liquidity. If prediction markets were like this, it would be better to rename them 'information bounty platforms.'

Overall, I believe that platforms like Polymarket, built on-chain, with transparent movements and anonymity, fulfill our expectations of decentralization—anyone can freely send signals through prediction markets without worrying about anything, only needing to be responsible for their own bets. The harsh reality is that there are too many wealth transfers caused by information asymmetry in the world, and those that occur in prediction markets or stock markets might be the ones we feel most keenly.

As for whether prediction markets themselves should maintain a relatively fairer environment for retail investors and whether they need to sustain a more acceptable positioning for the public for sustainable development, that is a question they should consider.

Conclusion

There are actually many interesting questions, such as whether there will be a prediction market in the future that can challenge Kalshi and Polymarket? If prediction market compliance continues to advance worldwide, will each country and region have its own leader (because these markets can do better localization and provide prediction events more friendly to users in their respective countries and regions), etc.

Prediction markets are not exactly new, but the compliance of prediction markets is a very new, very avant-garde bold attempt. I believe that over time, the various questions mentioned in this article will have interesting answers that we cannot anticipate.