Chapter 16: Further Reading - NBA Modeling

Academic Papers and Research

  1. Oliver, Dean. "Four Factors of Basketball Success." The foundational research identifying the four most important statistical categories in basketball. Originally published in the early 2000s, it remains the core framework for NBA analytics and team evaluation.

  2. Kubatko, Justin, Dean Oliver, Kevin Pelton, and Dan Rosenbaum. "A Starting Point for Analyzing Basketball Statistics." Journal of Quantitative Analysis in Sports (2007). A comprehensive review of basketball analytics methods and metrics, providing a roadmap for quantitative analysis of the sport.

  3. Manner, Hans. "Modeling and Forecasting the Outcomes of NBA Basketball Games." Journal of Quantitative Analysis in Sports (2016). Rigorous statistical analysis of NBA game forecasting models, comparing various approaches and evaluating their predictive accuracy.

  4. Lopez, Michael J. and Gregory J. Matthews. "Building an NCAA Men's Basketball Predictive Model and Quantifying Its Success." Journal of Quantitative Analysis in Sports (2015). While focused on NCAA basketball, the methodological insights apply directly to NBA modeling.

  5. Wolfson, Julian, Joseph S. Koopmeiners, and John P. DiNardo. "Who's the Best? Ranking and Rating in the NBA." Statistical Science (2021). Advanced treatment of ranking and rating methods applied to NBA teams and players, with careful attention to uncertainty quantification.

Books

  1. Oliver, Dean. "Basketball on Paper: Rules and Tools for Performance Analysis." Potomac Books (2004). The definitive book on basketball analytics. Introduces the Four Factors framework, discusses possession-based analysis, and provides the intellectual foundation for modern NBA quantitative analysis.

  2. Hollinger, John. "Pro Basketball Forecast." Annual publication that popularized PER (Player Efficiency Rating) and introduced many casual fans to basketball analytics. While PER has limitations, Hollinger's work was instrumental in mainstreaming NBA statistics.

  3. Silver, Nate. "The Signal and the Noise." Penguin Press (2012). While not basketball-specific, Silver's discussion of prediction and forecasting includes relevant NBA examples and provides essential context for thinking about uncertainty in sports prediction.

Data Sources

  1. NBA.com/stats. The official source for NBA statistics, including box scores, play-by-play data, player tracking data, and advanced metrics. The API (accessible through Python packages like nba_api) provides programmatic access to current and historical data.

  2. Basketball Reference (basketball-reference.com). Comprehensive historical NBA statistics including game logs, player advanced stats, team ratings, and draft data. An essential resource for historical analysis and cross-referencing.

  3. Cleaning the Glass (cleaningtheglass.com). Premium analytics site that provides NBA statistics filtered to remove garbage time, end-of-quarter heaves, and other noise. Their filtered metrics are more predictive than raw statistics for many purposes.

  4. PBPStats (pbpstats.com). Detailed play-by-play analysis including lineup data, on/off court splits, and possession-level statistics. Valuable for player impact estimation and lineup analysis.

  5. nba_api (Python package). Open-source Python package providing access to the NBA.com stats API. Enables programmatic data collection for box scores, play-by-play, tracking data, and more.

Online Resources

  1. Thinking Basketball (YouTube and blog). Ben Taylor's deep analyses of NBA players and teams using a combination of statistical and film analysis. His work on player impact metrics and historical player evaluation is particularly valuable.

  2. The BBall Index. Advanced player impact metrics and team projections using sophisticated statistical models. Their player impact estimates are among the most robust publicly available.

  3. Seth Partnow's Work (formerly The Athletic). Partnow's writing on NBA analytics, particularly on tracking data interpretation and the limitations of various metrics, provides essential context for modelers.

  4. Nylon Calculus (archive). A collaborative basketball analytics blog that produced many foundational articles on NBA statistical analysis before being absorbed into larger publications.

Betting Market Resources

  1. Haralabos Voulgaris' Public Discussions. Though not published in traditional form, Voulgaris' discussions of NBA betting strategy (in podcasts and interviews) provide rare insight from one of the most successful NBA sports bettors in history.

  2. Pinnacle Sports NBA Betting Resources. Pinnacle's editorial content on NBA betting covers market efficiency, live betting strategies, and bankroll management specific to the NBA's high-volume schedule.

  3. Action Network NBA Coverage. Provides line movement data, public betting percentages, and sharp action indicators for NBA games. Useful for understanding how the market is pricing specific games.

Methodological References

  1. Engelmann, Jordan. "Possession-Based Player Performance Analysis in Basketball." MIT Sloan Sports Analytics Conference (2017). Introduces PIPM (Player Impact Plus-Minus) and discusses the challenges of isolating individual player impact from team and lineup effects.

  2. Sill, Joseph. "Improved NBA Adjusted +/- Using Regularization and Out-of-Sample Testing." MIT Sloan Sports Analytics Conference (2010). Technical paper on using ridge regression and regularization to produce more stable adjusted plus-minus ratings for NBA players.