Prediction markets are open markets that use financial incentives to forecast specific outcomes. These markets are set up for betting on various event outcomes, with market prices reflecting the public’s view on the likelihood of an event occurring.

A typical prediction market contract trades between 0% and 100%. The most common form is a binary options market, where the contract settles at either 0% or 100% at expiry. Users can also sell options before the event occurs, exiting at the market price.

Through prediction markets, we can extract public expectations about future outcomes based on the value placed by participants on an event. Traders with different beliefs reflect their confidence in possible outcomes through contract trading, with the market prices of these contracts representing aggregated beliefs.

Prediction markets have a long history, almost as old as gambling itself, and their intersection with politics seems ancient: people in the medieval period were already enthusiastic about betting on the outcome of Catholic papal elections.

With the upcoming U.S. presidential election, interest in political betting surged in July due to events such as Trump’s assassination attempt, Biden’s withdrawal, and the Democratic Party’s replacement of Harris as a candidate. Prediction markets like Polymarket have gained widespread attention.

Polymarket: Order Book-Based Prediction Market

@Polymarket is a decentralized prediction market project launched in 2020, founded by Shayne Coplan and supported by renowned institutions and angel investors, including Polychain Capital, Founders Fund, and Vitalik.

Polymarket allows users to trade on highly controversial topics around the world (such as politics, sports, pop culture, and more), enabling them to build portfolios based on their predictions.

Unlike traditional sports betting, Polymarket allows users to freely trade shares on market topics before they conclude, providing speculators with flexible opportunities to engage in probability-based wagers.

Polymarket utilizes the Gnosis-based Conditional Tokens Framework (CTF). For every $1 worth of ERC-20 tokens, such as USDC, staked, two conditional tokens are issued. These tokens represent the binary outcomes of a trading event (e.g., whether an event will happen or not, yes or no). The market aggregates multiple binary outcome markets into a composite result.

The conditional tokens fluctuate in the market based on trading demand, and users can buy or sell them through the order book. Alternatively, once the event reaches a conclusion, holders of the correct token receive the entire $1 payout.

Since the two tokens trade independently in a market similar to a CEX, there may be instances where the combined price of the two tokens does not equal $1. This is where market makers come in to help balance the price discrepancies. Therefore, before the event concludes, you can always redeem the tokens for the $1 stake using a combination of the two tokens.

Polymarket’s prediction markets consist of the following key components:

• Market Themes: Each prediction market on Polymarket focuses on a specific theme or event. While users can submit new market creation proposals via Polymarket’s Discord, the platform retains discretion over which markets are created due to the complexity of wording.
• Oracles: Determining the outcome of an event typically requires input from an oracle. Polymarket uses the UMA optimistic oracle, which allows anyone to submit solutions. If no challenges are made to a solution within a given period, the solution is accepted as the fact. In rare cases of disputes, the oracle’s decision is resolved by UMA token holders.
• Conditional Tokens: As previously mentioned, locking $1 results in the issuance of two conditional tokens: “Yes” and “No.” At market settlement, the holder of the winning outcome receives the entire $1. The “Yes” and “No” tokens are freely traded in the market, with their prices reflecting probabilities. Polymarket uses a Conditional Tokens Framework (CTF) developed by the Gnosis protocol, based on the ERC1155 token standard.
• Order Book Market: Polymarket employs a hybrid on-chain order book trading mechanism, similar to dYdX v3. Users authorize trades via signatures, with operators matching orders off-chain, and the final interactions occur on-chain. The contract handles non-custodial settlement, performing atomic swaps between binary outcome tokens and collateral assets, so operators do not hold the $1 stake.
• Liquidity Providers: Unlike sports betting, Polymarket allows free trading of conditional tokens before the outcome is determined. Pricing is dictated by supply and demand rather than mechanisms, which may cause token prices to deviate (e.g., the combined price of the two tokens may not equal $1). Anyone can place limit orders to profit from the bid-ask spread, and Polymarket also offers additional incentives in USDC.

Polymarket currently has no plans for token issuance and does not actively incentivize users through a rewards program. Nevertheless, Polymarket has distributed over $3 million in USDC through its liquidity rewards program this year to incentivize market-making activities, aiming to enhance the overall liquidity depth of the platform. The highest volume markets currently pay approximately $600 USDC daily to liquidity providers.

SX Bet: Single-Bet Prediction Platform

@SX_Bet is a sports betting platform based on Ethereum, founded in 2019, and now operates on the SX Chain established on Arbitrum Orbit Rollup.

Currently, SX Bet primarily supports betting markets focused on sports topics, including major events in tennis, football, baseball, and basketball. Recently, new betting categories have been added, including Crypto, Degen Crypto, and Politics, with bets related to the price trends of major crypto assets and on-chain meme coins, as well as outcomes of the U.S. presidential election.

Unlike Polymarket, SX Bet follows a traditional sports betting model, supporting only single bets and not allowing for free trading of bets before the outcome of the predicted event is determined.

SX Bet’s innovation lies in its implementation of a combination betting system. Users predict a series of events, and only if all predictions are correct do they receive the payout. Combination bets often offer substantial rewards, serving as a form of leverage in prediction markets. SX Bet’s market makers act as the counterparties for these trades.

These combination bets resemble lotteries and can offer returns of up to thousands of times the stake, with their success stories often going viral. This is one of the most exciting aspects of traditional sports prediction markets.

Clearly, Polymarket and all prediction markets based on the “dual-token” conditional framework cannot offer combination bets, as contracts cannot mint a conditional token for every possible outcome combination and ensure it can be freely traded with sufficient liquidity. Markets with only two outcomes have limited odds and may be less attractive to users.

Pred X: AI-Driven Topic Prediction Market

@PredX_AI is a prediction market initially based on the Sei blockchain, covering a range of topics including politics, cryptocurrency price forecasts, and popular events. Currently, the platform supports betting with USDC across multiple blockchains such as Base, Linea, Sei, and Bitlayer, and has launched a corresponding Telegram mini-app. The Pred X Telegram mini-app, called PredXFun (@PredxFantasyBot), offers two modes: one is a game mode where users predict the probability of trending events and earn points; the other is a real mode where users connect their wallets to place bets on similar topics on the official website.

Unlike Polymarket, where prediction market topics are primarily proposed by users on Discord, Pred X’s prediction topics are mostly generated by Aimelia AI, which scrapes popular news and market sentiment indexes from the internet. These topics are then pushed to the Pred X website, where users spontaneously form trading markets. Although Pred X supports multiple blockchains, it is not a fully decentralized prediction market application. The prices for different outcomes of prediction topics are determined by the platform’s centralized order book, while the ordering process and market for each prediction topic are managed according to smart contract rules.

Objectively, Pred X is still an immature platform compared to other prediction markets. The order book depth and trading volume for predictable topics on the site are significantly lower than those of Polymarket and SX Bet. As a prediction market, it should ideally allow users to freely trade different outcome tokens before the event results are revealed. Unfortunately, the order book on Pred X does not support user-created orders. In the absence of market makers in most markets, users are effectively unable to freely trade outcome tokens. Furthermore, the documentation does not provide detailed explanations on how consistency between topic markets across different chains is maintained or how liquidity for all probability outcome tokens is ensured across chains. In the “real mode” of the Telegram mini-app, there is a discrepancy in betting prices for the same topic compared to the official website.

These issues cast doubt on the practical usability and reliability of Pred X. Overall, the product currently seems more like a work in progress.

Azuro: a betting protocol powered by liquidity pools

@azuroprotocol is not a prediction market itself but a foundational protocol for creating on-chain prediction markets. This comprehensive, permissionless infrastructure includes on-chain smart contracts and web components, enabling the creation of multiple prediction market applications based on Azuro. All Azuro-based betting platforms can be found at https://azuro.org/ecosystem.

Azuro supports only single bets and does not allow for the free trading of “yes” and “no” outcomes like Polymarket. Users can only receive payouts after the result is announced.

The Azuro system is centered around liquidity pools. Anyone can deploy their own liquidity pool by interacting with the Azuro factory contract. Multiple betting platforms can be created within a single liquidity pool, and each platform can offer various events for betting on different prediction topics.

In binary split models like Polymarket, liquidity is isolated and divided among various prediction events. Azuro introduces a concept known as the liquidity tree, where multiple events under a single prediction topic, or even multiple topics and platforms, can share the same liquidity pool.

The liquidity tree provides a hierarchical structure, with different potential events defining the liquidity range. For example, a football match has multiple possible score outcomes. This liquidity ensures that the platform can act as a counterparty for bets and pay out potential winnings (which may result in losses for liquidity providers). If bettors generally incur losses, liquidity providers can earn profits. A liquidity tree provides liquidity for multiple prediction topics and serves as a counterparty, generating profits or losses.

The odds for each event on Azuro are calculated based on the amount of funds bet on each event relative to the total liquidity of the prediction topic. Initial odds are set by specific data providers, who also provide initial liquidity. Data providers can adjust odds during the betting process, and the solvency of these odds is guaranteed by the initial liquidity.

Azuro also supports the implementation of multiple dApps. Betting platforms can set their own transaction fees, which bettors can choose freely, while liquidity pool creators can set the pool’s profit-sharing ratio. A portion of all pool profits is directed to Azuro’s own DAO, and Azuro has also issued its native token, $AZUR.

Conclusion

The philosophy behind prediction markets is fascinating. Participants aim to profit and view the free market as the most effective information-gathering system, enabling them to predict real-world events. These predictions often turn out surprisingly accurate. In an era where recommendation algorithms monopolize information, prediction markets seem to effectively uncover truths and reflect diverse viewpoints, as demonstrated by political predictions on Polymarket.

For many crypto users, their first encounter with prediction markets might have been through the Trump vs. Biden index launched on FTX during the last presidential election. Coupled with SBF’s strong market-making capabilities, it even allowed for high-leverage trading. Although centralized, it provided a highly engaging experience.

Cryptocurrencies have significantly reduced transaction friction in prediction markets, offering a better and more efficient market mechanism. Moreover, the concepts of smart contracts and AMMs have brought improved market mechanisms to prediction markets—providing permissionless access and better liquidity. Many AI AgentFi projects also view prediction markets as a battlefield for harnessing collective intelligence and honing their models.

However, there are notable drawbacks: While Polymarket allows for free trading of conditional tokens, it struggles with flexible betting mechanisms and lacks high-return expectations, which may diminish the enjoyment for some ordinary users. Azuro’s liquidity pool model is still somewhat complex and lacks post-bet trading capabilities.

Rather than focusing on the mechanisms and technological innovations, the current popularity of prediction markets should be seen as another instance of mass adoption in crypto culture. It represents a triumph of the free market culture behind it, which is especially valuable in an age where algorithmic authority increasingly dominates information. After all, nothing is smarter than the market, and no information system is more effective than a free market.

References

https://learn.polymarket.com

https://messari.io/report/yes-or-no-on-polymarket…

https://docs.polymarket.com

https://legacy-docs.polymarket.com/polymarket-+-uand…

https://sx.bet

https://docs.sx.technology

https://help.sx.bet/en/articles/6233471-parlay-betting-rules…

https://predx.ai

https://predxai.medium.com

https://azuro.org/#build

https://gem.azuro.org/concepts

https://gem.azuro.org/hub

Disclaimer：

1. This article is reprinted from [X], All copyrights belong to the original author [Trustless Labs]. If there are objections to this reprint, please contact the Gate Learn team, and they will handle it promptly.
2. Liability Disclaimer: The views and opinions expressed in this article are solely those of the author and do not constitute any investment advice.
3. Translations of the article into other languages are done by the Gate Learn team. Unless mentioned, copying, distributing, or plagiarizing the translated articles is prohibited.