Execution, not accuracy, drives profits

A decomposition of 222 million prediction‑market trades finds returns split into a directional (forecast) component and an execution (price) component; traders who are often right still lose money because they pay worse prices, while near‑random traders profit by securing better execution. Automated traders earn a persistent execution edge (about 2.52 cents per contract) that explains the profit gap. — This reframes how we interpret prediction markets: accuracy in forecasts does not guarantee financial reward, concentrating profits with automation and raising questions about access, market design, and the use of markets as public‑interest forecasting tools.

Sources

He called it the Pocket Crystal

Isegoria 2026.05.10 80% relevant

The article’s account of Marc Porat and General Magic (the 'Pocket Crystal') is a concrete case of the broader claim that having a visionary design/idea isn’t sufficient — market outcomes hinge on choices about constraints, execution, partnership and governance (Apple took board seats, partners, funding, but General Magic’s freewheeling engineer culture failed to ship), which maps onto the idea that execution matters more than theoretical correctness.

Who profits from prediction markets?

Tyler Cowen 2026.03.17 100% relevant

Joshua Della Vedova’s analysis of 222 million trades showing automated traders pay 2.52 cents less per contract and that forecasting skill and execution are nearly orthogonal.