Chapter 42 Further Reading: Research Frontiers

Causal Inference

  1. Pearl, J. (2009). Causality: Models, Reasoning, and Inference (2nd ed.). Cambridge University Press. The foundational text on causal inference using structural causal models and DAGs. Pearl's do-calculus provides the mathematical framework for determining when causal effects can be identified from observational data. Essential for any researcher serious about moving beyond correlational analysis in sports.

  2. Angrist, J. D., & Pischke, J.-S. (2009). Mostly Harmless Econometrics. Princeton University Press. A practical and accessible guide to causal inference in social science. The chapters on instrumental variables, regression discontinuity, and difference-in-differences are directly applicable to sports analytics. The emphasis on "credible" identification strategies is a valuable corrective to casual causal claims.

  3. Cunningham, S. (2021). Causal Inference: The Mixtape. Yale University Press. A modern, accessible textbook on causal inference with code examples. Covers DAGs, matching, IV, RDD, and difference-in-differences with an engaging writing style. Available free online, making it an excellent starting point for self-study.

  4. Imbens, G. W., & Rubin, D. B. (2015). Causal Inference for Statistics, Social, and Biomedical Sciences. Cambridge University Press. A comprehensive treatment of the potential outcomes framework for causal inference. The chapters on treatment assignment mechanisms and matching methods complement Pearl's graphical approach and provide additional tools for sports analytics researchers.

  5. Hernan, M. A., & Robins, J. M. (2020). Causal Inference: What If. Chapman & Hall/CRC. A rigorous textbook connecting causal inference theory to practical application. The treatment of time-varying confounding and sequential treatment strategies is particularly relevant to the sequential nature of betting decisions.

Reinforcement Learning

  1. Sutton, R. S., & Barto, A. G. (2018). Reinforcement Learning: An Introduction (2nd ed.). MIT Press. The definitive textbook on reinforcement learning. Covers the theoretical foundations (MDPs, dynamic programming, Monte Carlo methods, temporal-difference learning) and modern deep RL approaches. The multi-armed bandit chapters provide the theoretical basis for the market selection problems discussed in Section 42.3.

  2. Szepesvari, C. (2010). Algorithms for Reinforcement Learning. Morgan & Claypool. A concise, mathematically rigorous treatment of RL algorithms. Excellent for readers who want a deeper understanding of convergence properties, regret bounds, and the theoretical guarantees behind Thompson Sampling and other bandit algorithms.

  3. Russo, D. J., Van Roy, B., Kazerouni, A., Osband, I., & Wen, Z. (2018). "A Tutorial on Thompson Sampling." Foundations and Trends in Machine Learning, 11(1), 1-96. The most comprehensive tutorial on Thompson Sampling available. Covers theoretical properties, practical implementation, and extensions to complex settings. Directly relevant to the market selection and model selection applications in Chapter 42.

  4. Mnih, V., et al. (2015). "Human-Level Control Through Deep Reinforcement Learning." Nature, 518(7540), 529-533. The landmark paper on deep Q-networks (DQN) that ignited the deep RL revolution. While the application (Atari games) is different from betting, the architectural ideas and training techniques are transferable to complex betting environments.

Market Microstructure

  1. Kyle, A. S. (1985). "Continuous Auctions and Insider Trading." Econometrica, 53(6), 1315-1335. The seminal theoretical model of informed trading and price formation. Kyle's lambda --- the price impact of order flow --- is a foundational concept for understanding how betting lines move and how information is incorporated into prices. Essential reading for any microstructure analysis.

  2. Glosten, L. R., & Milgrom, P. R. (1985). "Bid, Ask and Transaction Prices in a Specialist Market with Heterogeneously Informed Traders." Journal of Financial Economics, 14(1), 71-100. The foundational model of bid-ask spreads as a function of adverse selection. The insight that spreads compensate market makers for the risk of trading against informed participants translates directly to the vig decomposition discussed in Section 42.4.

  3. Easley, D., & O'Hara, M. (1987). "Price, Trade Size, and Information in Securities Markets." Journal of Financial Economics, 19(1), 69-90. Introduces the sequential trade model that leads to the PIN (Probability of Informed Trading) metric. Adapted to betting markets, PIN can estimate the fraction of sharp betting activity in a given market.

  4. O'Hara, M. (1995). Market Microstructure Theory. Blackwell. A comprehensive textbook on market microstructure covering inventory models, information models, and market design. While focused on financial markets, the frameworks for analyzing price formation, information asymmetry, and market maker behavior apply directly to betting markets.

Sports Betting Efficiency and Strategy

  1. Levitt, S. D. (2004). "Why Are Gambling Markets Organised So Differently from Financial Markets?" Economic Journal, 114(495), 223-246. A landmark paper showing that sportsbooks set lines to maximize profit by exploiting bettor biases, not to balance action. This challenges the standard efficient markets interpretation and provides a theoretical foundation for understanding market mispricing.

  2. Fama, E. F. (1970). "Efficient Capital Markets: A Review of Theory and Empirical Work." Journal of Finance, 25(2), 383-417. The original formulation of the efficient market hypothesis. Understanding the three forms of efficiency (weak, semi-strong, strong) provides the framework for assessing which types of information are and are not reflected in betting prices.

  3. Dixon, M. J., & Coles, S. G. (1997). "Modelling Association Football Scores and Inefficiencies in the Football Betting Market." Journal of the Royal Statistical Society: Series C, 46(2), 265-280. A classic paper demonstrating a model-based approach to finding inefficiencies in soccer betting markets. The bivariate Poisson model for soccer scores remains influential, and the paper's approach to identifying market mispricings is a template for quantitative betting research.

  4. Kaunitz, L., Zhong, S., & Kreber, J. (2017). "Beating the Bookies with Their Own Numbers." arXiv:1710.02824. Demonstrates that a strategy based on odds discrepancies across bookmakers can generate systematic profits. The paper also documents the account limitation response from sportsbooks, providing empirical evidence for the optimal account management problem discussed in Section 42.1.

Emerging Methodologies

  1. Chernozhukov, V., et al. (2018). "Double/Debiased Machine Learning for Treatment and Structural Parameters." Econometrics Journal, 21(1), C1-C68. The foundational paper on double machine learning (DML), which combines machine learning flexibility with causal inference guarantees. DML enables causal effect estimation in high-dimensional settings, making it applicable to sports analytics where many potential features and confounders exist.

  2. Athey, S., & Imbens, G. W. (2019). "Machine Learning Methods That Economists Should Know About." Annual Review of Economics, 11, 685-725. A survey connecting machine learning to econometric (causal) goals. Covers causal forests, heterogeneous treatment effects, and the intersection of ML with causal inference. Provides a roadmap for applying these methods to sports data.

  3. Vovk, V., Gammerman, A., & Shafer, G. (2005). Algorithmic Learning in a Random World. Springer. The theoretical foundation for conformal prediction. Conformal methods provide distribution-free prediction intervals with guaranteed coverage, directly useful for uncertainty quantification in sports prediction and bet sizing.