A perspective on prediction markets: centralized versus permissionless protocols

Prediction markets, once niche experiments, have evolved into important financial instruments. These platforms, where participants deliberate on the outcomes of future events, have attracted significant attention due to their proven ability to be more accurate than polls and traditional commentators, especially regarding critical political and economic outcomes. Its rise is fueled more by individuals’ desire to exploit their knowledge for profit and a broader cultural obsession with instant data and future results, resulting in hundreds of millions, and sometimes billions, of dollars flowing into these markets every week.

The success of the industry creates a demand for billions of dollars. The current landscape is mainly formed by duopolies, Calici and Polymarket. These two platforms, although seen in direct competition, represent two different approaches to the same market. Calici is signed as a regulated exchange, while Polymarket is the decentralized cryptocurrency market. A new competitor, Rain, has recently emerged, built with a distinct and unrestricted architecture that aims to address the structural limitations of existing entities.

This comparison is made between these three distinct platforms, Calici, Polymarket, and Rain, focusing on four key areas: scalability and liquidity, resolving outcomes and trust, user experience and accessibility, and the fundamental tension between decentralization and centralization.

Sponsored

Sponsored

Central constraint: market creation liquidity

While the forecast market typically focuses on metrics like trading volume and active users, the real barrier to meaningful growth is a structural bottleneck known as “market-making liquidity.” This refers to the speed, cost and accessibility for any user to create a new business market. The current mainstream models Kalichi and Polymarket operate under a “publisher” model, acting as gatekeepers, limiting their ability to scale completely.

Kalshi: regulatory bottleneck

Calici’s position in the market is defined by its approach that places importance on regulatory compliance. As a centralized platform based in the United States, it is fully regulated by the CFTC as a dedicated contract market. This regulatory clarity gives access to traditional financial institutions, institutional hedgers and individual users who rely on cash and certainty of value.

However, this regulatory framework imposes a “regulatory bottleneck”. insertion process Types New markets are a long legal function, not just a technique, because their model is essentially authorized by regulators. A notable example is the CFTC’s initial rejection of Calici’s proposal for contracts based on elections, considering them “gambling”, which led to a costly lawsuit against its own regulator to eventually list the markets.

As a result, Kalichi is structurally limited to the inclusion of a small number of large-scale mass events, or the “major” part of the demand curve. Its focus is limited to markets that are profitable enough to justify massive legal and lobbying costs, such as major sports or economic data. The growth of the platform is clearly stifled by the pace of the judicial system, as it fights with ongoing legal battles for its sports contracts in different states of the United States. Its market-making liquidity is almost non-existent, as it is licensed by law.

Polymarket: The human bottleneck

Polymarket, the decentralized perspective, represents the largest global marketplace for digital asset forecasts. It is known for its transparency on the chain, self-incubating funds, and generating a massive volume in political, cultural and digital events.

Despite its decentralized branding and on-chain mechanics, Polymarket is structurally built as a “permissioned service”, not a permissionless protocol. Its official documentation claims that markets are created by its internal team with input from the community, revealing a “human barrier”. Its success depends on its editorial judgment, and it operates more like a media company.

Sponsored

Sponsored

This model is not radically scalable; The expansion of the number of markets requires a relative expansion of the editorial team. While an impressive volume (38,270 new markets in a peak month) is generated by a centralized team, it is only a statistical fraction of the capabilities of a completely user-generated and unauthorized system. Polymarket’s market-making liquidity is low and regulated, as it is authorized by a team.

Rain: A platform approach that should not be allowed

Designed with scale in mind across AMM and cross-chain tools, Rain is a new protocol designed specifically to solve the “Market Making Liquidity Crisis”. Its architecture represents a shift from a “publisher” model to a true “platform” model.

The distinguishing feature of Rain is its allowed tool: any user can create a market. This aims to capture the “long tail of possibilities”, a concept where the total value of millions of products with limited demand is equal to the value of some “strikes”. While others discuss the “head” (such as presidential elections and major sporting events), “Rain” targets the infinite universe of private events ​​that are of interest to specific communities or companies, such as specific project deadlines, GitHub issues, or internal DAO votes. The value of the platform must be derived from the total trading volume of millions of private markets that are impossible to create on existing platforms.

This architecture also offers two distinct types of markets: public markets (visible to everyone) and private markets (which require an access code). Private market capacity has been classified as a new product category, turning prediction markets into a tool for coordinating active businesses. For example, a CEO could create a subsidized financial incentive market for the engineering team when a product needs to be shipped, a B2B market that Kalshi and Polymarket cannot serve.

Sponsored

Sponsored

Trust and resolve the results

Outcome resolution, the mechanism for determining a real-world outcome, is the most important confidence factor for prediction markets.

Center Bit (Kalshi)

Kalshi relies on traditional and centralized arbitrage, in accordance with stock exchange rules and regulatory oversight. Its internal team, subject to CFTC rules, serves as a “central arbitrator” or identifier. This method offers clarity, speed and potential for legal challenge to users.

The main risk, however, is a catastrophic “single point of failure”. The authority over the final decision rests with the operator and its regulatory counterparts. This is not only a technical risk, but an existential political risk, since the authority of the platform is delegated by the CFTC and could be revoked by a new political administration or by a judicial ruling, which could lead to a capital freeze. For institutional users, this compromise is often acceptable, but for others, it raises concerns that a central entity is being abused. Additionally, this model leverages human interaction with platform limitations and is not scalable to the “long tail” of markets.

Decentralized oracles (Polymarket)

Polymarket uses blockchain transparency, decentralized oracles and conflict protocols to make results auditable. Its basic solution mechanism is based on optimistic hexes (UMA), a “trust by default” model where an answer is proposed and assumed to be correct unless challenged. This system reduced ambiguity, but required a robust hex design and was vulnerable to manipulation in low liquidity scenarios.

Sponsored

Sponsored

A notable incident exposed a vulnerability where an attacker with a large amount of $UMA tokens successfully manipulated a government vote to force an incorrect result. This incident reveals a conflict of interest as the token holders (voters) could also be market participants (bettors). In response, the UMA is moving to a new model that involves abandoning open solutions and creating a “white list of experienced proposals” that in effect re-centralizes the resolution mechanism. The move trades government attack directives for a new risk of centralization and collusion.

AI Hybrid (RIN)

The Rain model aims to combine transparency and speed by eliminating human gatekeepers. This model advances fair results that is based on artificial intelligence to obtain additional transparency while maintaining decentralization. The system focuses on automated chain solutions supported by algorithms, and is a consensus system for many artificial intelligence models.

Rain’s multi-stage system is designed for scalability and security.

• Initial solution. For public markets, the creator or inventor AI can be chosen as the initial solution. The invention of artificial intelligence aims to provide low-cost, unbiased, data-driven results. For private markets, the creator solves the result (for example, a CEO who solves the internal market of the company).
• Mechanism of conflict. After the initial resolution, the “Dispute Window” will open. Each participant can make a dispute by offering financial guarantees, an economic bet that prevents arbitrariness. An AI judge investigates the dispute and can change the solution. If the losing side escalates the dispute further, it is examined by “decentralized human oracles” to make a final and binding decision.

This architecture provides a scalable and automated way to solve millions of public market “long tails” via AI oracles. The conflict system also functions as a motivated economic predicate, similar to the optimistic system, but using a more robust decentralized human predicate, rather than token voting which has been shown to be manipulable.

Conclusions

The prediction market industry has been validated by the “old guard” of Kalshi and Polymarket, causing multi-billion dollar demand while simultaneously exposing their structural ceilings. They operate as services and publishers, restricted by legal and human barriers respectively. The 1000x growth opportunity in this sector will not be discovered by fighting the same “core” markets. Instead, it lies in the open innovation of the “long tail of possibility.” The real value is not in predicting a presidential election onebut rather in forecasting the ten million project deadlines, supply chain results, and community voting processes that make up the new “long tail” of our economy. Achieving this future requires a protocol built on three pillars: open creation, scalable solutions via mechanisms such as AI-powered oracles, and long-tail native properties such as private markets. The development of this space represents a transition beyond being just another place of business, it makes predictions themselves platformised.