The Evolution of Prediction Markets

A Castle Research Report: Diving into Polymarket and Azuro

The future is inherently uncertain, and throughout history, humans have always gambled trying to predict it, whether through monetary wagers or strategic actions.

Is it going to rain tomorrow?

Is my enemy going to attack today?

Getting those predictions right was a matter of life or death for a long time.

Historically, most of these markets have been “closed,” either centralized or managed by third parties. This means participants must adhere to predefined rules without the ability to create their own markets or set their own conditions.

The advent of cryptocurrency has presented an opportunity to shift these markets on-chain, making them genuinely decentralized and resistant to manipulation. 

Unlike centralized markets, where users trade against the "house", decentralized prediction markets allow users to trade against others who take the opposite side of their bets. 

Since the launch of Augur in 2015, prediction markets have emerged as a prominent use case for cryptocurrencies, even being mentioned in the Ethereum whitepaper. 

Despite this, they have only just recently captured mainstream attention.

Prediction markets are experiencing unprecedented growth across all metrics, finding product-market fit (PMF) amid speculation surrounding events such as the launch of the Ethereum ETF and the outcome of the US elections. 

This report aims to provide a comprehensive overview of the prediction market landscape, introducing the concept, highlighting historical use cases, examining the current state of the art, and delving into two leading platforms, Azuro and Polymarket, exploring the future potential of this sector.

We think it’s going to be a winner - wanna bet?

Between Gambling and Cryptocurrency: The Significance of Prediction Markets

Prediction markets are one of the most compelling use cases for crypto, enabling users to place bets on a wide range of events on-chain.

Ancient Greek gamblers depicted on a vase

Gambling has been ingrained in human culture since the Paleolithic period. 
By 3000 BCE, the Sumerians were already using six-faced dice for betting. 

Over the millennia, humans have bet on just about anything.

The proliferation of technology and the rise of digital natives have shifted gambling towards online platforms, making online gambling a booming industry. 

Source: Casinosenligne

  1. Worsening economic situation globally

  2. The rise of digital native users

  3. The convenience and accessibility of online gambling

  4. Increased Regulatory Clarity 

The Decentralized Casino Thesis

Nearly a third of the global population lives in countries where gambling is either in a legal grey area or outright banned and persecuted. Officially, these restrictions are intended to protect citizens from the risks of gambling. However, they end up not solving the problem, driving individuals to unlicensed operators

As a result, users face worse odds, higher fees, and a significantly increased risk of falling victim to fraud or scams perpetrated by unscrupulous operators.

To fill this gap, in just a few years, global online platform giants such as Rollbit and Stake have risen to be among the most popular online gambling platforms sponsoring highly influential celebrities, sporting events, and teams – and guess what? They both accept crypto. 

Within this context, prediction markets are particularly well positioned at the intersection between crypto and gambling, fitting into the “decentralized casino thesis” brought forward by GCR.

They have much less friction than other crypto use cases, as they don’t require users to learn all the nuances about how the system works under the hood.

Users are already familiar with online betting, making the transition to prediction markets seamless and requiring minimal education or changes in habits.

On Prediction Markets

Before delving into the significance of prediction markets, it is essential to introduce the concept. Prediction markets are platforms where users can bet on the outcome of future events. 

They function similarly to futures markets but cover a broader range of topics beyond asset prices, including sports, politics, weather, and even the likelihood of your favorite crypto influencer going to jail. 

Under the hood, prediction markets operate through smart contracts that encode the rules of each market, ensuring a fair and transparent system. 

Users participate by buying shares valued at either 0 (for incorrect predictions) or 100 (for correct predictions) based on the outcome. 

How Do Prediction Markets Work?

  • Users interact with the dApp using stablecoins.

  • They bet on their preferred outcome for an event

  • Based on the outcome, they either receive a payout or lose their wager

Prediction markets can either leverage an Order Book (Polymarket) or an Automated Market Maker, or AMM, design (Azuro).

In order book-based prediction markets:

  • Users can choose market orders or limit orders

  • The liquidity of markets is directly correlated to user activity and interest

  • The order book provides transparency on the current Bids and Asks in each market 

In an AMM design:

  • Users are incentivized to deposit liquidity in pools, forming a “unified betting platform” across all markets in exchange for a percentage of the trading fees.

  • The AMM adjusts odds and defines prices and other conditions inside the prediction engine.

  • A Prediction engine determines all the logic of accepting bets, payout distribution, and more. 

There are various types of markets based on their outcomes:

  • Binary: YES/NO (e.g., is BTC hitting 100k next month?)

  • Categorical: multiple options (e.g., next US president)

  • Scalar: range-based options (e.g., what will the BTC price be next month?) 

History of Prediction Markets and Early Use Cases

The Augur Protocol

Prediction markets are not a new primitive to the world of cryptocurrency.

They were among the first use cases, with the Augur protocol and its token, $REP, being one of the earliest examples.

Augur is a decentralized oracle network for prediction markets. Its Initial Coin Offering (ICO) was in 2015, and its V1 launched three years later, in 2018. 

Source: dappradar

Augur aimed to solve a critical issue in blockchains: how to transfer real-world data to blockchains without relying on third parties, known as the “oracle problem.” This solution enabled users to bet on a wide range of markets, including controversial ones like murder.

In 2020, Augur V2 was launched with several improvements:

  • A Peer-to-peer order-book design

  • Trading using DAI as the base currency

Why did Augur Fail? 

Despite these enhancements to its product, Augur struggled to gain traction, as shown by the data below.

Several factors contributed to this outcome:

  • Early Entry: When Augur launched, Uniswap V1 had yet to be deployed, and many current mechanisms for providing liquidity were nonexistent. Those were the days of Etherdelta. Prediction markets on the platform also struggled to find a solid PMF, often experiencing a decline in usage after initial excitement.

  • Accessibility: There was no easy way to access the platform. Users had to download the Augur application client and have a functioning Metamask Wallet with an ETH balance. 

  • Complexity: The platform’s mechanisms and user interface were complicated, leading to network problems and lengthy connection times.

  • Limited market amounts: A restricted number of available markets limited user engagement.

  • High-gas Cost: Built on Ethereum, Augur transactions were subject to ETH gas costs.

The Prediction Market Sector

State of the Art

Fast forward four years to 2024, and the prediction market sector is experiencing unprecedented interest and growth. 

The US elections have acted as a significant catalyst for the prediction markets sector, reflecting the growing interconnectedness between cryptocurrency and real-world events, such as ETF approvals and US candidates' political stances on Bitcoin.

Remarkably, a US presidential candidate referenced decentralized prediction markets for election data, highlighting the sector's increasing relevance.

Prediction markets are now recognized as one of the most accurate ways to gauge genuine public sentiment on various topics. This increased attention has translated into rising volumes and user interest. 

Next, let's explore the current state of prediction markets using on-chain data and the leading platforms, Azuro and Polymarket, as case studies.

Understanding Prediction Markets Using Data 

The prediction market sector holds a Total Value Locked (TVL) of $162m.

While the Defillama chart shows the sector since its inception, most TVL accumulation occurred in 2024.

Source: Defillama

Similarly, volumes have surged from a maximum of $25 million weekly before June 2024 to a weekly volume of $183 million in August 2024, more than a 7x increase.

As a result of their recent growth, approximately 30,000 users engage with prediction markets daily, marking a significant increase from the previous average of 3,000 users experienced until May 2024.

This surge is consistent with the rise in daily volumes, as most new volumes stem from users who joined Polymarket to bet on major events like the US elections.

The Case of Azuro and Polymarket

This section analyses two of the leading players in the sector, Azuro and Polymarket

It provides a deep dive into how these platforms work under the hood, their differences, and their similarities, arguing that rather than being competitive, these platforms are complementary, each with its nuances.

Polymarket

Polymarket has become a household name in the prediction market sector since its launch in 2020. By many, Polymarket is heralded as the Trojan Horse that finally brought crypto into the mainstream. 

Based on Polygon, Polymarket accounts for $87.5 million in TVL, representing over 60% of the sector’s total TVL.

Polymarket has gone through somewhat of an evolution in its own right. Initially, markets were deployed with a “constant product” AMM, similar to Uniswap V2 pools, where the prices of the shares of each market reflect the probabilities of the outcomes. However, over the last year, a shift towards an orderbook model has occurred.

As it stands, each market takes the form of an order book, where users can share their Bids and Asks.

Polymarket leverages UMA’s decentralized Optimistic Oracle for dispute resolutions, with UMA token holders being impartial judges on the outcomes of the relevant markets. In a disagreement on the result of an event, “disputers” can challenge the answer. 

Since its inception, Polymarket has had:

  • Over $1.4b in volume

  • 3.1m transactions

  • Over 150k total bettors on its platform

As of August 2024, Polymarket’s most popular categories by Bet Volume are predominantly political, except for one sports Category related to the Euro 2024 winner.

Polymarket appears to satisfy larger-sized, sporadic major event wagerers, instead of catering to small customers who bet often. 

The two most popular categories on Polymarket, “Presidential Election Winner 2024” and “Democratic Nominee 2024,” have contributed over $660m or 84.23% of the total volume.

As we approach the 2024 US Presidential Elections, it will be interesting to observe whether the cumulative traded volume trend will be similar to that of 2020.

While Polymarket's spike in volumes can be attributed to the US elections, all the platform’s on-chain metrics have been trending positively since the beginning of 2024 and the launch of the Bitcoin ETF, indicating a broader trend of increasing adoption of prediction markets.

Future Plans

As part of its roadmap, Polymarket is committed to increasing the number of supported markets and reducing its reliance on Political events. 

Recent events such as the Olympics have contributed to these markets picking up, with the prediction market for the most gold medals at the Paris Olympics becoming the third biggest market, with over $6m in volume.

Furthermore, they just open-sourced PolyLend, their p2p lending protocol, which allows users to “borrow USDC against conditional token positions” (their collateral locked in existing markets).  

PolyLend offers:

  • Leverage by borrowing USDC against conditional token positions

  • Fixed-interest rate loans on USDC, overcollateralized by conditional token positions

  • Dutch auction mechanisms for transferring loans to new lenders

Interestingly, Polymarket said this development would not be used in production but rather open-sourced for the community to build on top of.

Azuro

Azuro is an infrastructure and liquidity layer for gaming dApps and protocols to build upon, currently live on Polygon, Gnosis Chain, and Chiliz.

Azuro introduces a peer-to-pool design, solving the inefficiencies of the peer-to-peer model, which were:

  • Dependency on active liquidity

  • Need for market markets to ensure user engagement

  • Activity concentrated on significant events

Under this new model, market makers and users interact by providing liquidity in a pool that serves multiple markets. In this way, anyone can act as the house and provide liquidity against bettors, creating “passive liquidity” and real yield for liquidity providers, who act as the counterparty of bettors on Azuro.

Adding more markets on top of a single pool also contributes to diversifying the risk of liquidity providers and improving the capital efficiency of the platform, allowing it to scale:

  • Supporting more volumes

  • Enhancing the user experience

  • Increasing yield returns for liquidity providers.

The design leverages a novel data structure that tracks provided liquidity, called the Liquidity Tree

Azuro’s liquidity pools serve as counterparties for trades on the platform, allowing users to provide liquidity and effectively become the house, generating real yield from the fees collected by the protocol. 

This way, users providing liquidity to a pool get immediate exposure to all the markets supported by that pool and earn fees based on the spread and volume, which are “embedded” in the markets' odds.

Under Azuro, dispute resolutions are managed by an “optimistic oracle approach,” which they plan to decentralize in the near future. As Azuro mainly supports sports events, they have clear-cut outcomes and are rarely subject to ambiguous situations such as political or news-related events. The AzuroDAO will act as the last resort arbiter of truth in the event of a dispute over an oracle's resolution outcome (implementation soon).

Aside from supporting sports markets and more games, Azuro allows anyone to launch a prediction application or website much faster and with no upfront costs required, such as bookmaker.xyz, a prediction application.

These frontends can abstract all the core tech away, avoid bootstrapping liquidity, and tap directly into the liquidity of Azuro, all while earning a percentage of the profits realized by the pool associated with their users' activity, similar to an affiliate model. 

Since its launch, Azuro has facilitated:

  • Over $200m in prediction volume

  • Over 5.8m prediction-related transactions

  • Over 30k total bettors

Azuro’s user base has been growing positively, reaching an all-time high of over 2,500 new users. It is also interesting to observe how most Azuro users are returning users.

Why is this the case? 

Only 0.7% of Azuro's volumes come from their most significant event, so their bet volumes are widespread. On top of this, they offer a wide range of events and games to keep their users engaged.

For this reason, Azuro primarily attracts users who prefer recurring bets on various markets, focusing on sports.

Let’s dive deeper into which categories are driving the majority of volumes on Azuro, as of August 2024:

  • 58.4% Football events

  • 11.9% Basketball

  • 11.2% Tennis

  • 9.6% Baseball

The only non-sport events within Azuro's top events are eSports games: Counter-Strike (placing 6th), League of Legends, Dota 2, and CS:GO.

Within these categories, what are the markets that drive the most volume?

All 10 of Azuro's most successful events have been football matches. English Premier League matches take the lead in single bet volume, with the Arsenal-Chelsea fixture in April 2024 reaching almost $1.5m. 

Regarding the number of bettors, international football matches are the most popular.

Since its inception, Azuro has generated over $4m in revenue, distributed to liquidity providers, with an average APY of over 20% (12% currently).

In the case of Azuro, liquidity providers face counterparty risk as the liquidity pools act as a counterparty to all trades on the platform and its affiliates.

The $AZUR Token

Azuro’s native token $AZUR, is used as the governance token and to add layers of gamification and incentive alignment between users and the protocol.

The existence of a native token for Azuro is one of the main differentiation factors to Polymarket.

For a complete overview of $AZUR, refer to Azuro’s technical documentation.

Future Plans

As part of its roadmap, Azuro strives to support a broader range of events, including social and political events: they recently announced the US Presidential Elections winner market, the first non-sport market for the protocol.

 Aside from expanding markets, Azuro is putting efforts into:

  • The decentralization of its governance and dispute-resolution mechanism

  • Supporting more networks with cross-chain liquidity

  • Launching no-coding front-ends

  • Further value accrual mechanisms for $AZUR and the DAO

Comparative Analysis

Firstly, we can see the differences in the technical design of each prediction market protocol, demonstrated below.

Augur was ahead of its time, just like Etherdelta, building a traditional orderbook product on a chain that hadn’t yet scaled or reached adequate performance. Polymarket initially utilised Uniswap’s success in the constant product formula to enable a greater user experience, before transitioning to an orderbook, much like Verterx Finally, Azuro implemented a Peer-to-Pool model, much like GMX, this enabled Azuro to scale liquidity across all markets evenly with deep pools that serve multiple markets, abstracting complexity away from the user. 

Aside from the technicals, we can also make business comparisons: 

  • Political vs. Sports Markets:
    Before the US elections narrative, Azuro was the sector leader due to its offering of sports-related markets. However, US elections have been the main driver of the recent success of prediction markets, captured primarily by Polymarket. At the same time, Azuro volumes have slowed in the summer due to their reliance on sports events, which have a break period during these months.

  • One-off vs. Recurring Events:
    While Polymarket thrives on large but few markets, attracting strategic narrative speculators, they most likely are using Polymarket as a one-off, whereby they have specifically come to a market to place a wager. Azuro on the other hand, provides many smaller markets, enticing users to return, converting them into recurring customers. The top event on Polymarket accounts for 62% of the total volumes, while for Azuro, this is only 0.74%. This highlights how Polymarket positioned itself to capture the most significant “news” events, compared to Azuro’s offer, which comprises several smaller events and is more suited for recurrent betting.

  • Token or no token:
    Whether or not Azuro’s token has much impact on their volumes and user growth, having a token provides a broader range of strategies to align the incentives between the protocol and the community.

  • B2B vs B2C: 
    The platforms are complementary rather than competitive. While Azuro focuses more on B2B and recurring sports events, Polymarket focuses more on B2C and news, political, and economic events. 

  • Growth, growth, growth: 
    Both platforms are looking to expand the roaster of markets supported as well as expanding to more networks.

Challenges for Mainstream Adoption

As novel tools, prediction markets face several challenges before they can be adopted beyond crypto:

  • Liquidity: 
    Low participation often results in poor market depth, leading to wider bid-ask spreads and less accurate predictions. This low liquidity can also make markets more susceptible to price manipulation by a few participants.

  • Regulatory Uncertainty:
    The legal status of prediction markets varies by jurisdiction, and the lack of clear regulations creates significant legal risks. Compliance with laws like AML and KYC is complex, particularly for decentralized platforms that emphasize anonymity.

  • Market Manipulation: The open, pseudonymous nature of these platforms and the relatively small size of some markets make them vulnerable to manipulation.

  • User Experience: 
    The complexity of decentralized platforms, including the need to manage wallets and understand gas fees, can be a barrier for non-technical users, limiting broader adoption.

  • Oracles and Data Accuracy:
    Prediction markets rely on oracles to bring external data onto the blockchain. Ensuring the reliability and security of these oracles is crucial, as inaccurate data can lead to false outcomes and loss of trust.

  • Scalability:
    High transaction costs and slow processing speeds on some blockchains can make participation expensive and less appealing, especially during periods of network congestion.

  • Incentives and Governance: 
    Properly aligning incentives for participants, market makers, and data providers is challenging. Decentralized governance can be slow and contentious, leading to inefficiencies in managing the platform.

  • Trust and Adoption: 
    Building trust in a decentralized environment without a central authority is difficult. Overcoming these trust issues and convincing users to transition from traditional platforms are critical for wider adoption.

Addressing these challenges requires ongoing innovation, regulatory engagement, and community development to ensure the long-term success of decentralized prediction markets.

Future of the Sector

In prediction markets, people put their money where their mouth is.

For this reason, they are increasingly recognized for their potential to provide accurate public sentiment on various topics.

Platforms like Azuro and Polymarket lead this sector, each with distinct strengths and market focuses. Azuro excels in sports markets and recurring events, while Polymarket dominates major political and news-related events.

As a short-term objective, both platforms want to expand their market offerings.

Polymarket aims to reduce its reliance on US elections, while Azuro wants to support more political and news markets and drive more volumes besides sports betting.

The future of prediction markets looks promising, with potential developments such as AI integrations and expanded market offerings enhancing their utility and appeal. 

One such speculation concerns AI agents, where prediction markets will move from manual market creation to AI-led.  For example, Azuro will use Olas to develop intelligent AI agents who can carry out specific tasks on its platform. In prediction markets, we can envision AI agents providing liquidity to, and drawing API data from markets efficiently.

Furthermore, they will take on the system's "guardians" role, solving conflicts and preventing disputed disputes, strengthening prediction markets' resilience and smooth operations.

As the sector evolves, prediction markets are poised to become a crucial tool for gauging public sentiment and making informed decisions in a decentralized manner.

In this way, using prediction markets can lead to truly decentralized sentiment, which can be integrated natively within news websites and mainstream newspapers and even used by politicians.

Prediction markets like Azuro and Polymarket may cement themselves even more as “real sentiment,” thus providing API services to news and data providers.

In fact, prediction markets are not only restricted to crypto users but are also interesting avenues for traditional gamblers, political experts, news publications, and others. 

This could also slightly change the platforms, expanding into decentralized surveys and other forms of polls to gauge public opinion, build theses, and test hypotheses and assumptions.

“While news noise continues to increase, prediction markets represent an “economically backed source of truth in a world with a growing problem of fake news, post-truths, and deep fakes.” 

Prediction markets have finally broken into the mainstream narrative.

Will they manage to sustain their positioning beyond seasonal hype? 

Are we expecting more challengers to come into this market once it solidifies? 

Food for thought.

⚔️ A Call to Arms

Do you have what it takes to enter the Castle and contribute to research, community initiatives, due diligence analysis, and advising/servicing projects in the space?

Or maybe you want to upskill and shadow community members who have already walked a successful path as an intern?

Brought to you by Francesco with the help of Atomist and Mooms.

Special thanks to Crypto Koryo for the Dune dashboards.

Thanks for reading, please give us a follow on Twitter at @Castle__Cap and visit our temporary website to learn more about our services and get in touch.

In our blog, we may discuss projects or tokens in which we hold positions. While we aim to provide informative content, our views are not financial advice. Please conduct your research and consult professionals before making investment decisions. Crypto markets are volatile, and past performance doesn't guarantee future results. Invest responsibly, and be aware of the risks. Your capital is at risk, and we do not accept liability for any losses.

Reply

or to participate.