Your name

James Grugett, Stephen Grugett, Austin Chen

Your email

[email protected]

Short one-sentence description of your proposed project

Prototype a new design for a decentralized prediction market.

Longer description of your proposed project Max 1500 words

We want to build a proof-of-concept prediction market that allows for greater decentralization and more participation—and thus more accurate predictions on more varied topics. Our design is based on:

• A dynamic pari-mutuel betting system,
• where individuals can create, promote, and resolve their own markets to earn a percentage of the transaction volume,
• through a smart contract on the Solana blockchain.

Today, prediction market creation is very centralized; e.g. the PredictIt team will decide which questions to add, often going for questions with wide appeal and interest like “Who will win the 2024 Democratic Nomination?”. This draws more traders to participate, but the information gained isn’t as novel, actionable or useful.

Augur (https://www.augur.net/) has tried allowing anyone to create a market with a complex system for reporting outcomes by votes from owners of their token. This implementation doesn’t obviously lead to the correct market resolutions. Voting on truth, in our view, does not accomplish the goal of true decentralization because a majority can always censor the minority. This problem—combined with Ethereum’s very high network fees—has essentially killed Augur.

Our idea to achieve better decentralization is to simply allow individuals to create and resolve their own markets however they like. In the long run, people will want to participate in markets focused on interesting topics where the creator has a good track record of deciding the outcome fairly.

By giving the market creator a cut (say, 1%) of each transaction, we encourage them to pose and resolve interesting questions, and promote their markets to their audience. We think that this model could lead to an explosion of niche markets curated by domain experts in different fields.

Motivating example: A Substack writer could create a prediction market for “Will the FDA approve X drug by Jun 10 2022?”

• Traders would trust the market creator based on his existing reputation, and place bets if they felt they had particular information they could profit from.
• The market creator receives a small portion of transaction volume, and thus has an incentive to publicize this market, thereby also improving market liquidity.

Q: Why Dynamic Pari-Mutuel betting?

There are two main ways to implement prediction markets: with a traditional order book to coordinate buyers and sellers or with an automated market maker (AMM) of some variety. We think an AMM is more advantageous for low-liquidity markets because it’s not required for someone to take the other side of each bet.