Prediction Markets: A Casino for Everything?

Foreword

Looking ahead to February 2026, if we were to assess which areas of the Web3 industry have already entered the mainstream, stablecoins and prediction markets would likely be the only two answers.

Stablecoins, needless to say, have become a key focus for governments, central banks, top financial institutions, and Fintech companies worldwide. Major events in 2025 include Circle's IPO, Stripe's acquisition of the stablecoin platform Bridge for approximately $1.1 billion, Trump's signing of the GENIUS Act, Visa, Mastercard, Citibank, and Coinbase's bidding war for stablecoin technology company BVNK, and Hong Kong's push for stablecoin regulations. The total transaction volume of stablecoins is expected to exceed $40 trillion in 2025, with a total market capitalization exceeding $300 billion. It can be said that stablecoins are becoming a conduit and pillar of the new financial system.

Are prediction markets casinos?

Both Polymarket and Kalshi claim that "event contracts" are compliant financial derivatives and are seeking full regulatory approval (such as from the CFTC).

Kalshi only serves US users, while Polymarket, although a global marketplace, currently restricts access to 33 countries, according to Polymarket's geo-blocking FAQ. Most jurisdictions restricting Polymarket, such as the EU, France, Poland, Singapore, and Portugal, classify it as a form of gambling, requiring specific licenses, which are often incompatible with Polymarket's decentralized, natively encrypted business model. Even in some US states, such as New York, there are proposed bans on prediction markets to avoid competition with state-owned exclusive betting licensees. According to the "Prediction Markets Report" issued by Keyrock in conjunction with platforms such as Dune, specifically regarding prediction market betting events, in 2025, Kalshi was primarily focused on sports, with sports accounting for approximately 85% of its trading volume. In contrast, Polymarket exhibits a more diversified mix—sports (39%), politics (34%), and cryptocurrency (18%), but sports still dominate. This highly overlaps with the sports betting industry. Analyzing users further, while Polymarket's users are predominantly crypto enthusiasts, a recent study, "Gambling on Crypto Tokens?", provides strong evidence of a link between crypto assets and gambling behavior. Researchers used Google Trends to proxy retail investor attention, revealing several significant patterns: Regions with high per capita lottery sales (DMA) in the US showed significantly higher Google search activity related to ICOs, indicating a high degree of overlap in behavior between gamblers and crypto retail investors; in regions with high per capita lottery sales, wallet activity surged rapidly after an ICO or NFT launch; and when gambling became legal in a state, crypto-related attention significantly decreased in high-lottery-sales regions. This study also indirectly suggests that users on Polymarket currently have a stronger tendency towards gambling. Meme coin players on the Solana chain (P-Jiang) are generally considered to have a stronger gambling tendency. In 2025, nearly 99% of Meme coins issued on the Pump.fun platform had gone to zero, and Solana's transaction fee revenue also dropped by more than 90%. The decline of Meme coins coincides with the rise of prediction markets. Applying this to our research, we can reasonably suspect that former P-Jiang players are becoming new users of Polymarket. Moreover, compared to speculating on Meme coins, prediction markets are more transparent and fair in their mechanisms. When P-Jiang buys event contracts, the final win or loss depends on the objective outcome of the event, without the risk of project-driven manipulation. In conclusion, although there is controversy regarding whether prediction markets are casinos in terms of regulation, due to the high degree of overlap between prediction content and the sports betting industry, and the fact that users are more inclined to gamble, the author believes that prediction markets at the current juncture can be seen as a form of "duck typing" for casinos (betting companies). "If it looks like a duck and quacks like a duck, then it must be a duck." In fact, prediction markets have advantages over traditional casinos: First, events in prediction markets are tradable during the event (early payouts/profit-taking); second, in casinos, the bookmaker unilaterally sets the odds and is the only counterparty, while in prediction markets, the counterparties are other users, so the platform doesn't bear the risk of gambling against them; finally, under the mechanism of prediction markets, even very niche long-tail bets can be successfully launched, such as "betting on whether it will rain in Beijing tomorrow." Regarding the last point, the potential of prediction markets clearly extends beyond the scope of "gambling." Prediction markets can create markets for hedging financial and macroeconomic events, such as allowing small and medium-sized business owners (like farmers and importers/exporters) to hedge against "supply chain risks" (such as whether a port will close) or "policy risks" (such as tariff changes). The Problem with Prediction Markets – Market Making Difficulty As an exchange, to ensure smooth trading, the exchange must guarantee the availability of contracts at all times. This requires a stable source of liquidity – market makers. Market makers profit from the bid-ask spread, that is, buying low and selling high. In the conventional cryptocurrency trading market, if the timeframe is extended, mainstream assets will still exhibit a certain degree of fluctuation, and the upward and downward trends often rotate in cycles. Therefore, market makers can continuously make money in the cryptocurrency trading market. The trading instruments in prediction markets are essentially event contracts. Each contract has a specific settlement time, and the formula "Yes + No = 1" means that only one contract will ultimately have a value of 1 USDC, while the others will become zero. This means that bets in prediction markets will eventually end rapidly in an extreme one-sided market trend from a certain point in time, leaving market makers with very little time to react. To consistently make money, market makers need to capture profit opportunities as much as possible and avoid inventory risk when one-sided market trends occur. Market making is far more difficult than in the conventional cryptocurrency trading market. The above discussion about the difficulty of market making only applies to events with huge trading volumes (US elections, high-profile sporting events). Niche long-tail trading has almost no activity and exhibits extremely wide bid-ask spreads, making it almost impossible for market makers to profit in such situations. Furthermore, some events involve numerous insiders with close ties to or who are themselves sources of information. Market makers are naturally at an informational disadvantage against these players, and the liquidity they provide becomes a channel for these insiders to cash out, inevitably leading to losses for market makers. The difficulty of market making means limitations in liquidity creation, which directly impacts the user's trading experience. To address this issue, Polymarket and Kalshi, in addition to returning a portion or even all of their commissions to market makers, have also invested heavily in liquidity subsidies to attract more market makers. Polymarket has invested approximately $10 million in liquidity subsidies, and Kalshi has invested at least $9 million. In fact, market-making subsidies have become a solid competitive advantage for leading platforms like Polymarket and Kalshi, which is why other prediction markets have struggled to succeed. Without market-making subsidies, market-making can only be sustained through transaction fee revenue, requiring increased trading volume and fee income, which the current size of prediction markets is far from sufficient. In conclusion, although many proponents define prediction markets as decentralized information aggregation and pricing systems, or "truth machines," current prediction markets are clearly overly focused on sports betting and short-term cryptocurrency price manipulation—a "dopamine casino"—lacking long-term social information value. The author is not against prediction markets; on the contrary, the author is a Polymarket user. The growth of prediction markets essentially stems from people's gambling demand on "events"—whether political events, sporting matches, or reality itself. This demand is real and enduring, and we don't need to cloak it in any way, much less label it as a "truth machine." In fact, many financial activities in human society originate from gambling, only with clearer regulations, more explicit ethics, and more rational designs. Prediction markets have the potential to become a more widespread financial tool; institutions use them to hedge against uncertainty, retail investors use them to reap dopamine, insider traders use them to sell the truth, and AI uses them for quantitative trading. Welcome to prediction markets!