I'm excited to share the litepaper for the Parimutuel Market Maker, a new market structure for permissionless prediction markets. I encourage you to visit the site linked below and read it in full, or bookmark it to read later. Here's a brief summary: The PMM is a parimutuel market at its base with dynamic pricing curves on top of each outcome. These curves dynamically change the cost of entering an outcome in the parimutuel market, and guarantee users a minimum return they'll receive if correct at the time of entry, a first for parimutuels. PMM markets can: 1. Bootstrap liquidity and odds for any market 2. Trade live during the event 3. Not require a subsidy to be liquid A parimutuel market is used as the base of the market to bootstrap liquidity and odds. This requires no sophisticated understanding of market structure, only the intuitive understanding of paying X to win Y if correct. There is no separation of liquidity providers and users. Every new buy simultaneously gives the user the directional exposure they want and provides liquidity to the market like a parimutuel, increasing the total capacity of the market. This means markets can start small and grow in capacity as they grow in popularity without needing to deposit money into a pool or onboard market makers. The dynamic curves on each outcome change the cost of buying into an outcome based on the relative demand for that outcome. If more users want to buy outcome A than outcome B, the price of outcome A will rise and B will fall. When a user buys an outcome, they are given a minimum return floor that guarantees their minimum payout if correct. This floor is only a minimum, and can grow throughout the market as more liquidity enters opposing outcomes. If a user buys in at 50% odds and holds through the outcome dropping to 25%, their minimum return will rise as new buyers increase the value of their claim on counterparty pools. If their outcome resolves as correct, they will receive a higher payout than the one they bought in at as a reward for providing liquidity through volatility. A user's payout is a combination of the directional exposure they have in an outcome and the rewards they accumulate by holding through volatility. Ratcheting returns incentivize liquidity to enter a market and hold through volatile periods. If a market has sufficient growth or volatility, a user who provided liquidity in both outcomes can see their payout for both rise and be profitable no matter which outcome occurs. This structure allows creators or liquidity providers to seed markets without being structurally disadvantaged. In testing, popular markets with volatile odds like short term price markets or sports can have a positive expected value for the earliest users, providing a clear path to liquid markets. Why not use an existing market structure? Orderbooks are the dominant market structure for prediction markets, and work well for popular events. The world cup has plentiful data on each team, making it possible for market makers to provide deep, liquid markets. As the data available for an event decreases, market makers become less able to confidently provide liquidity to the market. In an extreme example, a soccer game between friends, there is no data available about either team, making it impossible for an external market maker to provide liquidity. To begin a permissionless market, there must be a way for unsophisticated users to initialize odds and liquidity. Parimutuels bootstrap odds and liquidity with ease. Money can be placed in each outcome pool, and the ratio of dollars in each outcome sets the odds. Though parimutuels initialize markets well, they must be closed for trading before an event begins. The payout a user is promised is not locked until the market closes, as new money entering their outcome decreases their ownership over the counterparty pools. While a parimutuel can start a market, it cannot provide the dynamic trading experience required for prediction markets. AMMs solve liquidity for the long-tail of tokens by making it profitable to passively provide liquidity to tokens, but don't work in prediction markets. The LP in a prediction market will hold only valueless tokens at resolution, meaning the fees they collect must be greater than the LP itself to make the LPs whole again. This is fine if the market creator has an external incentive to make the market, like to aggregate information or for marketing, but is not sustainable for a creator-made market economy to flourish. The PMM allows creators to make markets of any topic while giving a structural advantage to being the first liquidity in the market. Markets can start with any amount of initial liquidity, large or small, and scale as the interest in the market grows. This gives creators the tools necessary to cheaply experiment with new market types to find what markets are most interesting to their audience. With the proper tools in place, permissionless markets will expand to serve every niche at the scale of the internet. No centralized team can match the breadth of topics nor out distribute a decentralized network of creators. In the age of the internet, decentralized creation has beat curation every time. Prediction markets are no exception. I encourage you to read the full litepaper. If you're interested to learn more or get an early look at the Melee platform, please dm me or drop a comment and I can reach out!