Chapter 2 Key Takeaways
Summary Card
Chapter Title: A Brief History of Prediction Markets Core Theme: Prediction markets have evolved from ancient betting traditions through academic experiments to modern, blockchain-powered platforms — driven by a consistent insight that markets aggregate dispersed information effectively, and constrained by recurring tensions between accuracy and regulatory legitimacy.
Historical Milestones at a Glance
| Year | Milestone | Why It Matters |
|---|---|---|
| ~1503 | Papal election betting documented in Rome | Earliest well-documented prediction market activity |
| 1688 | Lloyd's Coffee House opens | Insurance as collective risk prediction |
| 1860s-1940s | U.S. political betting markets flourish | Demonstrated accuracy long before modern platforms |
| 1945 | Hayek's "The Use of Knowledge in Society" | Theoretical foundation for markets as information processors |
| 1988 | Iowa Electronic Markets (IEM) founded | First modern academic prediction market |
| 1996 | Hollywood Stock Exchange (HSX) launches | Proved play-money markets can be accurate |
| 1999 | TradeSports founded in Dublin | First major real-money exchange model |
| 2001 | InTrade launches | Became the world's most prominent prediction market |
| 2003 | DARPA FutureMAP canceled | Showed political and communication risks; raised public awareness |
| 2004 | Wolfers & Zitzewitz publish "Prediction Markets" | Definitive academic survey of the field |
| 2005 | Tetlock's Expert Political Judgment | Showed experts are poor forecasters; motivated alternatives |
| 2008 | Arrow et al. "Promise of Prediction Markets" letter | Nobel laureates call for regulatory reform |
| 2011 | Good Judgment Project begins | Superforecasters outperform intelligence analysts |
| 2013 | InTrade closes | Governance failure, not market failure |
| 2014 | PredictIt receives CFTC no-action letter | New model for U.S. legal prediction markets |
| 2015 | Metaculus founded | Reputation-based forecasting without money |
| 2018 | Augur v1 launches on Ethereum | First decentralized prediction market |
| 2020 | Polymarket founded | Leading crypto-based prediction market |
| 2020 | Kalshi receives CFTC DCM designation | First fully regulated U.S. prediction market exchange |
| 2022 | Manifold Markets launches | Play-money innovation with frictionless market creation |
| 2024 | Court rules Kalshi can list election contracts | Landmark regulatory precedent |
Key Figures
| Person | Affiliation | Primary Contribution |
|---|---|---|
| Robin Hanson | George Mason University | LMSR, idea futures, futarchy, prediction market theory |
| Justin Wolfers | University of Michigan | Empirical analysis, influential surveys |
| Eric Zitzewitz | Dartmouth College | Empirical analysis, manipulation studies |
| Kenneth Arrow | Stanford University (Nobel laureate) | Advocated for regulatory reform |
| Charles Manski | Northwestern University | Influential critique (prices vs. probabilities) |
| Philip Tetlock | University of Pennsylvania | Superforecasting, Good Judgment Project |
| Robert Forsythe | University of Iowa | Co-founded the IEM |
| Forrest Nelson | University of Iowa | Co-founded the IEM |
| Joyce Berg | University of Iowa | IEM accuracy research |
| Thomas Rietz | University of Iowa | IEM accuracy research |
| John Delaney | InTrade/TradeSports | Founded the largest pre-blockchain prediction market |
| Shayne Coplan | Polymarket | Founded the leading crypto prediction market |
| Tarek Mansour | Kalshi | Co-founded the first CFTC-regulated prediction market |
Platform Comparison
| Platform | Type | Years Active | Key Innovation | Status |
|---|---|---|---|---|
| IEM | Real money (academic) | 1988-present | First modern prediction market | Active |
| HSX | Play money | 1996-present | Mass-audience play-money market | Active |
| InTrade | Real money (commercial) | 2001-2013 | Broad event contracts, media prominence | Closed |
| PredictIt | Real money (academic) | 2015-? | U.S. political markets via no-action letter | Winding down |
| Augur | Crypto (decentralized) | 2018-present | First blockchain prediction market | Active (v2) |
| Polymarket | Crypto (centralized UX) | 2020-present | Stablecoin settlement, high liquidity | Active |
| Kalshi | Real money (regulated) | 2021-present | Full CFTC DCM designation | Active |
| Metaculus | Reputation-based | 2015-present | No money; reputation and community incentives | Active |
| Manifold Markets | Play money | 2022-present | Frictionless market creation | Active |
Six Core Lessons from Prediction Market History
1. Markets Aggregate Information Effectively
From Roman grain markets to Polymarket, the core insight holds: financial incentives (or strong non-financial incentives) cause people to reveal honest assessments, and the resulting prices are informationally rich.
2. Accuracy Is Demonstrated but Not Perfect
The IEM and other markets have compiled impressive track records, generally outperforming polls and often matching sophisticated statistical models. But prediction markets are not oracles — they can be wrong, especially when outcomes depend on small groups (Supreme Court justices, conclave cardinals) or when liquidity is thin.
3. Regulation Is the Persistent Challenge
Every major prediction market has faced regulatory obstacles. The pattern is consistent: innovation outpaces regulation, platforms operate in gray zones, and regulatory action eventually follows. Success increasingly requires proactive regulatory engagement (Kalshi model) rather than regulatory avoidance (InTrade model).
4. Multiple Models Coexist and Compete
Real-money, crypto, play-money, and reputation-based platforms each serve different needs and different regulatory environments. No single model has dominated; the ecosystem is diverse.
5. Communication and Perception Matter
The FutureMAP cancellation demonstrated that technical merit is insufficient. Prediction markets must be explained and framed carefully to avoid triggering moral objections related to gambling on negative events.
6. Governance and Trust Are Non-Negotiable
InTrade's collapse showed that even accurate markets fail if the platform operator cannot be trusted. Customer protection, transparent finances, and succession planning are as important as contract design.
Vocabulary Review
| Term | Definition |
|---|---|
| Prediction market | A market where contracts pay off based on the outcome of future events; prices reflect collective probability estimates |
| No-action letter | A CFTC regulatory instrument stating the agency will not take enforcement action against a specified activity |
| DCM (Designated Contract Market) | A CFTC-regulated exchange authorized to list futures and options contracts |
| LMSR | Logarithmic Market Scoring Rule — an automated market maker designed by Robin Hanson |
| Event contract | A financial contract whose payoff depends on whether a specified event occurs |
| Favorite-longshot bias | The tendency for markets to overweight favorites and underweight longshots |
| Futarchy | Robin Hanson's proposal to use prediction markets for governance decisions |
| Superforecaster | A term from the Good Judgment Project for individuals with exceptional forecasting accuracy |
| Brier score | A scoring rule that measures the accuracy of probabilistic predictions (lower is better) |
| Calibration | The degree to which predicted probabilities match actual observed frequencies |
| Play-money market | A prediction market using virtual currency rather than real money |
| Order book | A list of buy and sell orders for a contract, organized by price |
| Double auction | A market mechanism where both buyers and sellers submit orders |
| Smart contract | Self-executing code on a blockchain that automatically enforces contract terms |
| Stablecoin | A cryptocurrency designed to maintain a stable value relative to a fiat currency (e.g., USDC) |
Connections to Other Chapters
- Chapter 1 introduced the concept of prediction markets; this chapter provided historical context.
- Chapter 3 will explain the mechanics (order books, market makers, pricing) that underlie the platforms discussed here.
- Chapter 4 on probability and calibration will formalize the accuracy analysis from the IEM case study.
- Chapter 5 on market design will draw on the LMSR and other mechanisms introduced in this chapter's discussion of Robin Hanson's work.
- Later chapters on strategy and trading will build on the understanding of platform differences established here.
This key takeaways card accompanies Chapter 2: A Brief History of Prediction Markets.