Challenges: Metrics

Challenges: Prediction Markets

Challenges: Decision Markets

PM aggregates valuable information when the decision rule is applied

Requirement: at the moment the decision rule is applied, the prices should have imported the largest quantity of private information possible. Also, prevent decision selection bias.

Options:

• modulate thickness of PM so it’s thickest before the measurement period, thinner during the measurement
  • either by changing the b parameter of the LMSR
  • either by adding/removing liquidity in a CPMM
• aggregate PM prices over the measurement period with a band-pass filter with smooth edges, to prevent gaming
  • also, in the CPMM case, it would maybe be possible to modulate the measurement based on thickness at any given point in time
• we’d better everyone know when these measurements are done to prevent randomization