blog: Don Marti


Some ways that bug futures markets differ from prediction markets

30 August 2017

Question about Bugmark: what's the difference between a futures market on software bugs and a prediction market? We don't know how much a bug futures market will tend to act like a prediction market, but here are a few guesses about how it may turn out differently.

Prediction markets tend to have a relatively small number of tradeable questions, with a large number of market participants on each side of each question. Each individual bug future is likely to have a small number of participants, at least on the "fixed" side.

Prediction markets typically have participants who are not in a position to influence the outcome. For example, The Good Judgment Project recruited regular people to trade on worldwide events. Bug futures are designed to attract participants who have special knowledge and ability to change an outcome.

Prediction markets are designed for gathering knowledge. Bug futures are for incentivizing tasks. A well-designed bug futures market will monetize haters by turning a "bet" that a project will fail into a payment that makes it more likely to succeed. If successful in this, the market will have this feature in common with Alex Tabarrok's Dominant Assurance Contract.

Prediction markets often implement conditional trading. Bug markets rely on the underlying bug tracker to maintain the dependency relationships among bugs, and trades on the market can reflect the strength of the connections among bugs as seen by the participants.