Prediction markets look simple on the surface: trade Yes and No, read the probability-like price. Under the hood, different market mechanisms determine how prices form, how liquidity appears, and what kind of trading behavior the system rewards.
The Two Big Families
Most prediction markets fall into one of these models:
- Order book markets: prices form from bids and asks placed by traders.
- Market maker markets: an algorithm quotes prices and adjusts them after each trade.
Both can work well. The best choice depends on liquidity, user base, and whether the platform can rely on external market makers.
Order Book Markets
Order books are the classic exchange model:
- Traders place limit orders (buy or sell at a chosen price).
- Market orders trade against the best available quotes.
- Spreads and depth reflect actual supply and demand.
Pros:
- Efficient when liquidity is high.
- Transparent price discovery.
Cons:
- Can be illiquid in small markets.
- Wide spreads and empty books in niche topics.
Market Maker Models
Market makers create continuous liquidity:
- You can usually trade at any time, even in small markets.
- The algorithm adjusts prices as inventory changes.
Pros:
- Always tradable.
- Better UX for long-tail markets.
Cons:
- Large trades can move price sharply.
- Someone pays for liquidity (explicitly or implicitly).
Multi-Outcome and Combinatorial Markets
Advanced markets include:
- Multi-outcome questions (more than Yes and No).
- Conditional markets (if A happens, what is the probability of B).
- Combinatorial markets (many related outcomes).
These designs can be more expressive, but they increase complexity and resolution risk.
Key Takeaways
- Mechanism choice determines liquidity, spreads, and trader experience.
- Order books shine in deep markets, market makers shine in long-tail markets.
- More expressive market designs increase complexity and dispute risk.