Chapter 2 Exercises

These exercises are organized into five parts of increasing complexity. Part A tests recall and comprehension. Parts B and C require analysis and comparison. Parts D and E involve programming and research.

Part A: Historical Foundations (Recall and Comprehension)

Exercise 2.1 — Matching Milestones to Dates

Match each event to its correct year. Write the letter of the correct year next to each event.

Exercise 2.2 — The Papal Election Markets

In 3-4 sentences, explain why medieval papal election betting markets are considered an important precursor to modern prediction markets. What specific features did they share with modern markets?

Exercise 2.3 — Roman Maritime Insurance

Describe how foenus nauticum (Roman sea loans) functioned as an implicit prediction mechanism. What probability was embedded in the interest rate, and how did market forces ensure reasonable calibration?

Exercise 2.4 — Decline of 19th-Century Political Betting

According to Rhode and Strumpf (2004), political betting markets in the United States were both large and accurate in the late 19th and early 20th centuries. Explain the two primary factors that led to their decline by the mid-20th century.

Exercise 2.5 — Lloyd's of London Principles

List three principles demonstrated by Lloyd's of London that later became central to prediction market theory. For each, explain the connection between the insurance market principle and its prediction market counterpart.

Part B: Platform Analysis (Application and Analysis)

Exercise 2.6 — IEM Design Choices

The Iowa Electronic Markets imposed a $500 maximum investment per participant. Discuss the trade-offs of this design choice. How does it affect: - Market liquidity? - The informativeness of prices? - The risk of manipulation? - Regulatory compliance?

Exercise 2.7 — Comparing Market Types

Fill in the following comparison table for four types of prediction markets. Use specific examples from the chapter.

Exercise 2.8 — The FutureMAP Framing Problem

Senator Ron Wyden described the Policy Analysis Market as a "federal betting parlor on atrocities and terrorism."

(a) In what specific ways was this characterization inaccurate? (b) What was the actual design of the PAM, and what types of contracts was it intended to offer? (c) Propose an alternative framing that the project's advocates could have used to preempt this criticism. Write a 2-3 sentence "elevator pitch" for the project that avoids triggering moral objections.

Exercise 2.9 — InTrade vs. Kalshi Regulatory Strategies

InTrade and Kalshi took fundamentally different approaches to U.S. regulation. Compare and contrast their strategies by addressing: - Legal jurisdiction - Regulatory engagement - Contract scope - Ultimate outcomes

Which approach has proven more sustainable, and why?

Exercise 2.10 — Play-Money Accuracy

The Hollywood Stock Exchange demonstrated that play-money markets can produce accurate forecasts. However, some researchers argue that play-money markets are inherently less reliable than real-money markets.

(a) What arguments support the claim that real money is necessary for accurate prediction markets? (b) What arguments suggest that play money can be sufficient? (c) What does the evidence from HSX, Manifold Markets, and Metaculus suggest?

Part C: Critical Thinking (Analysis and Evaluation)

Exercise 2.11 — The Manipulation Question

During the 2024 U.S. presidential election, a single trader on Polymarket (known as "Fredi9999") made large trades that visibly moved market prices.

(a) Does this demonstrate that prediction markets are vulnerable to manipulation? (b) Wolfers and Zitzewitz (2004) argued that manipulation attempts in prediction markets typically fail because other traders correct distorted prices. Under what conditions might this self-correction mechanism break down? (c) How should we distinguish between a well-informed trader moving prices toward the truth and a manipulator moving prices away from the truth?

Exercise 2.12 — Manski's Critique

Charles Manski (2006) argued that prediction market prices cannot, in general, be interpreted as probabilities because the mapping between prices and probabilities depends on participants' risk preferences.

(a) Explain Manski's argument in your own words. (b) Under what conditions would market prices closely approximate true probabilities? (c) How significant is this critique in practice? Does it undermine the practical usefulness of prediction markets?

Exercise 2.13 — The Morality of Prediction Markets

Prediction markets on negative events (pandemics, wars, terrorist attacks) face persistent moral objections.

(a) Articulate the strongest moral argument against allowing people to profit from predicting negative events. (b) Articulate the strongest counterargument. (c) Some platforms have addressed this concern by donating profits from "negative event" markets to charity. Evaluate this approach: does it resolve the moral tension?

Exercise 2.14 — Forecasting Without Markets

Metaculus and the Good Judgment Project demonstrate that accurate forecasting is possible without financial markets.

(a) What motivational structures do these platforms use to substitute for financial incentives? (b) In what situations might non-market forecasting methods outperform prediction markets? (c) Could prediction markets and non-market forecasting systems be combined? Propose a hybrid system and describe its potential advantages.

Exercise 2.15 — Regulatory Prediction

Based on the historical patterns described in Section 2.9.2, predict the likely regulatory trajectory for prediction markets in the United States over the next 5-10 years. Consider: - The Kalshi court ruling - The PredictIt litigation - The growth of blockchain-based platforms - Political attitudes toward prediction markets

Write a 200-300 word assessment.

Part D: Data Analysis and Programming

Exercise 2.16 — Timeline Enhancement

Starting with the code in code/example-01-timeline-visualization.py, add the following enhancements: 1. Add at least 10 additional events not included in the original dataset (research these yourself). 2. Add a "zoom" feature that shows only events within a specified date range. 3. Add a bar chart showing the number of events per decade.

Exercise 2.17 — IEM Accuracy Analysis

Using the synthetic data and code in code/example-02-iem-analysis.py: 1. Calculate the mean absolute error (MAE) for IEM predictions vs. actual outcomes for each election. 2. Calculate the same metric for polling averages. 3. Create a chart comparing the two, similar to those in Berg et al. (2008). 4. At what time horizon before the election do IEM predictions begin to outperform polls?

Exercise 2.18 — Platform Lifespan Analysis

Using the data in code/example-03-platform-comparison.py: 1. Calculate the average lifespan of prediction market platforms in each category (academic, commercial, blockchain, play-money). 2. Create a visualization showing platform lifespans on a timeline. 3. What is the survival rate of prediction market platforms launched before 2010 vs. after 2010?

Exercise 2.19 — Sentiment Analysis (Advanced)

The DARPA FutureMAP controversy generated extensive media coverage. Write a Python script that: 1. Creates a synthetic dataset of 50 newspaper headlines about FutureMAP (use realistic language for 2003-era coverage). 2. Classifies each headline as positive, negative, or neutral toward prediction markets. 3. Visualizes the distribution of sentiment. 4. Compares the sentiment in headlines published before vs. after the cancellation.

Note: You may use simple keyword-based classification rather than NLP models.

Exercise 2.20 — Historical Accuracy Tracker

Build a Python class called HistoricalAccuracyTracker that: 1. Accepts a list of prediction market forecasts (each with a predicted probability and an actual outcome). 2. Calculates Brier scores for the forecasts. 3. Generates a calibration plot. 4. Compares the calibration to a baseline (random forecasts or always-50%).

Test your class with synthetic data representing IEM election forecasts from 1988-2024.

Part E: Research and Synthesis

Exercise 2.21 — Annotated Timeline

Create a detailed annotated timeline of prediction market history from 1988 to the present. For each event, include: - Date - Brief description (1-2 sentences) - Significance (1-2 sentences explaining why this event matters) - Category (academic, commercial, regulatory, technological)

Your timeline should include at least 40 events.

Exercise 2.22 — Platform Post-Mortem

Choose one prediction market platform that has shut down (InTrade, Augur v1, a TradeSports, or another platform you research). Write a 500-word "post-mortem" analysis covering: - What the platform did well - What caused it to fail - What lessons modern platforms should learn from its experience

Exercise 2.23 — Country Comparison

Research the legal status of prediction markets in three different countries (e.g., United States, United Kingdom, Australia, or others). For each country: - Describe the relevant regulatory framework - List any prediction markets that operate or have operated there - Assess whether the regulatory environment is favorable or hostile to prediction markets

Present your findings in a comparison table with a 200-word synthesis.

Exercise 2.24 — The Academic Debate

Write a 400-word essay summarizing the academic debate between prediction market advocates (Hanson, Wolfers, Zitzewitz) and skeptics (Manski). Address: - What are the core claims of each side? - What evidence supports each position? - Where does the current consensus lie? - What questions remain unresolved?

Exercise 2.25 — Prediction Market for Prediction Markets

Design a prediction market contract that asks: "What will be the largest prediction market platform (by trading volume) in 2030?"

(a) Define the contract precisely — what exactly constitutes "trading volume" and how would you measure it? (b) What resolution criteria would you use? (c) What challenges would this contract face (liquidity, time horizon, definitional ambiguity)? (d) What would current prices likely look like if this contract existed today?

Exercise 2.26 — Technology Forecast

Based on the technology evolution described in Section 2.9.3, write a 300-word forecast of prediction market technology in 2030. Consider: - Blockchain scalability improvements - AI integration (automated market making, question generation) - Mobile and social media integration - Cross-platform liquidity aggregation

Exercise 2.27 — The Superforecaster Connection

Read the summary of Tetlock's Superforecasting (2015) or the key findings of the Good Judgment Project. Write a 300-word analysis of how the superforecasting approach relates to prediction markets. Specifically address: - What do superforecasters and prediction markets have in common? - Where do they differ? - Could they be combined, and how?

Exercise 2.28 — Historical Counterfactual

Consider the counterfactual: "What if the DARPA FutureMAP project had not been canceled in 2003?" Write a 300-word speculative analysis of how prediction market history might have unfolded differently. Consider: - Would government adoption have accelerated? - Would InTrade have faced less regulatory pressure? - Would blockchain-based markets have emerged earlier or later?

Submission Guidelines

• Part A: Brief written answers (1-4 sentences each, except where longer responses are specified).
• Part B: Written analyses of 100-200 words each unless otherwise specified.
• Part C: Essays of 200-400 words each unless otherwise specified.
• Part D: Submit Python code files (.py) with clear comments and output.
• Part E: Submit written reports in Markdown format.

These exercises accompany Chapter 2: A Brief History of Prediction Markets.