Circle Internet Financial’s research arm recently published a blog post that compares centralized and decentralized prediction markets platforms.

The blog first noted that the main idea behind prediction markets is to use “the wisdom of the crowd” – meaning that “individuals have unique biases, [however] gathering the knowledge and opinions of many individuals cancels out these individual biases, resulting in more robust and accurate predictions.”

Centralized Networks May Be “Unnecessarily Expensive”

Before the launch of decentralized networks, almost all prediction market platforms were “centrally owned and operated” – which makes them “unnecessarily expensive”, the blog noted.

This is due to the high network usage fees that range from 3% to 10% of profits.

According to Circle’s research team, centralized prediction markets are also prone to “manipulation” as they are usually controlled by a single entity. Moreover, the researchers point out that since regulatory policies vary from one jurisdiction to another, it can become difficult to use centrally managed prediction markets.

For instance, centralized prediction markets may be regulated as derivatives and options trading in one jurisdiction, and as gambling in another, the blog mentioned.

More “Incentive To Be Honest” On Centralized Platforms 

Proponents of decentralized prediction markets state that they “improve reporter accountability” because there is a “hard-coded protocol” that allows users to “challenge inaccurate reports” – while also penalizing those who “report incorrectly.”

However, malicious reporters on decentralized networks only risk losing their staked tokens, whereas, entities managing centralized platforms risk losing a lot more – such as their entire business, reputation, and financial losses, Circle’s blog noted.

The research report concluded: “the incentive to be honest is greater for a centralized entity than a reporter in decentralized networks.”

Additionally, the researchers analyzed several major decentralized prediction markets including Augur (REP), Stox (STX), and Gnosis (GNO).

Most Major Decentralized Prediction Markets Are Built On Ethereum 

As covered, Augur is a blockchain-based platform developed on the Ethereum blockchain, and Circle’s researchers have determined that Augur “currently has the most users and activity” – compared to all other decentralized prediction markets.

Stox has also been built on the Ethereum network and its native token, STX, is ERC-20 compliant. ”Unlike Augur and Gnosis, Stox employs a single token model”, the research report noted. Moreover, “Stox only allows trading outcomes of events that have a discrete number of potential outcomes” – meaning “markets must be binary or categorical and cannot be scalar or continuous.”

Gnosis is another major decentralized markets platform that has been built on Ethereum.

Gnosis Allows Developers To Launch More Apps On Top Of Base Layer  

Gnosis’ team and community aims to “provide third party developers with tools to build new prediction market applications on top of Gnosis because the team believes that different categories of prediction markets should have different trading interfaces and different marketing and regulatory strategies.”

As mentioned, Augur is the most active decentralized prediction markets platform and it “experienced a surge in bets during the 2018 midterm elections, surpassing $1.5 million”, the report pointed out.

DappRadar also reported that on November 7th, Augur recorded an all-time high in volume of 2935 ETH (appr. $600,000). According to the State of the Dapps, there are an average of 66 daily active users on Augur and about 44 daily users on Stox.

Both DappRadar and the State of the Dapps have not reported any data on Gnosis.