Skip to content
Longterm Wiki
Back

Market Scoring Rules

web
eecs.harvard.edu·eecs.harvard.edu/cs286r/courses/fall12/papers/HansonLogMS...

Hanson's MSR paper is foundational to prediction market design and has applications in AI safety for aggregating expert forecasts about AI risks, timelines, and policy outcomes; the URL appears inaccessible but the paper is widely available via Harvard CS course materials.

Metadata

Importance: 62/100conference paperprimary source

Summary

Robin Hanson's foundational paper introducing Market Scoring Rules (MSRs), a mechanism that combines proper scoring rules with prediction markets to create automated market makers that always accept trades. MSRs provide incentives for truthful probability reporting while maintaining liquidity without requiring matched counterparties, making them practical tools for information aggregation.

Key Points

  • •Market Scoring Rules combine proper scoring rules with prediction markets, creating a subsidized automated market maker that guarantees liquidity
  • •The mechanism incentivizes participants to update market probabilities toward their true beliefs by offering payment based on scoring rule improvements
  • •MSRs solve the thin-market problem in prediction markets by having a market maker always willing to take the other side of any trade
  • •The logarithmic market scoring rule (LMSR) became the dominant formulation, widely used in real-world prediction markets
  • •MSRs have bounded worst-case loss for the market operator, making them practical for real-world deployment in information elicitation

Cited by 1 page

PageTypeQuality
Prediction Markets (AI Forecasting)Approach56.0

Cached Content Preview

HTTP 200Fetched Mar 15, 20260 KB
Session Verification 

 
 
 
 
 
 
 
 
 
 

 
 
 JavaScript Required

 To protect our systems and ensure reliable access for all users, Archive-It uses a bot protection system that requires JavaScript.

 Please enable JavaScript in your browser and try again. Some browser extensions may block JavaScript, so disabling them may help.

 If you believe you're seeing this in error or need help, contact us at: aitratelimit (at) archive.org.
Resource ID: 1be83497a42bf07c | Stable ID: MGEzMjhmMm