On March 11, 2026, the renowned college basketball statistician and founder of KenPom, Ken Pomeroy, offered an insightful masterclass on predictive analytics in sports during his appearance on the Moneyball podcast, hosted by Knowledge@Wharton. The hour-long episode delved deep into the methodologies that underpin the acclaimed KenPom rankings, exploring the enduring power of tempo-free efficiency ratings, the "four factors" of basketball success, and the nuanced role of preseason priors in generating the most accurate forecasts for college basketball, particularly as the high-stakes March Madness tournament looms. This discussion highlighted how sophisticated statistical models continue to revolutionize understanding and prediction in sports, moving beyond conventional metrics to uncover the true drivers of team performance.
The Rise of KenPom: A Paradigm Shift in Basketball Analytics
Ken Pomeroy’s website, KenPom.com, has become an indispensable resource for college basketball enthusiasts, coaches, and professional handicappers alike since its inception in 2002. Frustrated by the limitations of traditional box scores and win-loss records in truly assessing team strength, Pomeroy, a former meteorologist with a deep understanding of statistical modeling, pioneered a system that aimed to quantify team efficiency on a per-possession basis. This approach fundamentally shifted the conversation around college basketball analysis, moving away from simple scoring averages and field goal percentages to a more granular, context-aware evaluation. His methodology quickly gained traction because it offered a more robust and objective measure of a team’s capabilities, independent of game tempo or opponent strength. The site’s daily updates and sophisticated algorithms have consistently provided some of the most accurate predictions in the sport, influencing everything from bracket pools to coaching strategies.
Pomeroy’s foundational premise, as discussed on Moneyball, revolves around the idea that basketball is ultimately a game of possessions. Traditional statistics often fail to account for the pace at which a team plays, making direct comparisons between high-tempo and slow-tempo teams misleading. By normalizing performance data to a per-possession basis, KenPom effectively creates an apples-to-apples comparison, revealing a team’s true offensive and defensive efficiency. This tempo-free approach was revolutionary, providing a clearer picture of which teams were genuinely better at converting possessions into points and preventing opponents from doing the same. It allowed analysts to discern skill and execution from mere pace, a critical distinction in a sport characterized by diverse coaching philosophies.
Decoding the "Four Factors" of Basketball Success
A cornerstone of the KenPom system, and a key topic of discussion during Pomeroy’s Moneyball appearance, is the concept of the "four factors." These factors, initially popularized by basketball analyst Dean Oliver in his book "Basketball on Paper," represent the most crucial statistical components that contribute to winning basketball games. Pomeroy adopted and refined these factors, integrating them into his efficiency calculations to provide a comprehensive statistical fingerprint for every team. The four factors are:
-
Effective Field Goal Percentage (eFG%): This metric adjusts standard field goal percentage to account for the fact that three-point shots are worth more than two-point shots. By giving 1.5 times the weight to three-pointers, eFG% provides a more accurate measure of shooting efficiency, reflecting a team’s ability to score efficiently from the field. A team with a high eFG% is effectively maximizing its shooting opportunities, regardless of whether those shots are from inside or beyond the arc. For example, a team shooting 50% from two-point range and 30% from three-point range might have a lower overall field goal percentage than a team shooting 48% from two-point range and 35% from three-point range, but the latter would have a higher eFG% due to the increased value of the three-point makes.
-
Turnover Percentage (TO%): This factor measures how often a team turns the ball over per possession. Turnovers are essentially "empty possessions" where a team fails to even attempt a shot, giving the opponent a free opportunity. A low turnover percentage indicates a disciplined offense that values possession and limits wasted opportunities, while a high percentage suggests sloppiness and gift-wrapped chances for the opposition. In a game of limited possessions, every turnover represents a significant loss of potential scoring, making this a critical determinant of success.
-
Offensive Rebounding Percentage (OR%): This metric quantifies the percentage of a team’s missed shots that it manages to retrieve. Offensive rebounds provide "second-chance" opportunities, extending possessions and offering additional scoring attempts. A high offensive rebounding percentage signals a team’s tenacity and ability to create extra scoring chances, which can be particularly impactful in close games or against stout defenses. While often overlooked, these second chances can significantly inflate a team’s scoring output over the course of a game.
-
Free Throw Rate (FTR): This factor measures a team’s ability to get to the free-throw line, calculated as free throw attempts divided by field goal attempts. Free throws are often considered "free points," and a high free-throw rate indicates an aggressive offense that attacks the basket, draws fouls, and converts those opportunities into guaranteed points. It also reflects a team’s ability to put pressure on the opposing defense, potentially getting key players into foul trouble. Teams that consistently get to the line and convert their free throws gain a significant advantage, especially in tight contests.
Pomeroy emphasized that these four factors collectively account for the vast majority of variance in basketball game outcomes. By analyzing a team’s performance across these dimensions, one can gain a far deeper understanding of its strengths and weaknesses than by simply looking at wins and losses or points scored. This granular, efficiency-based analysis forms the bedrock of KenPom’s predictive models.
The Dynamic Role of Preseason Priors
One of the more sophisticated aspects of KenPom’s methodology, and a point of particular interest for the Moneyball hosts, is the integration and evolution of "preseason priors." At the very beginning of a college basketball season, there is limited in-season data available to accurately rank teams. Therefore, KenPom, like many predictive models, relies on "priors" – educated guesses or initial estimates of team strength based on historical data, returning player talent, recruiting rankings, and coaching changes. These preseason priors serve as a starting point, providing a reasonable initial ranking when empirical game data is scarce.
As the season progresses and more games are played, the model gradually updates these priors with actual performance data. This is a classic example of Bayesian inference, where initial beliefs are refined and improved as new evidence becomes available. Pomeroy explained that the weight given to preseason priors diminishes steadily throughout the season, replaced by the increasingly robust in-season efficiency metrics. By mid-season, the priors have very little influence, with current performance dominating the rankings. However, their importance in the early weeks cannot be overstated, as they prevent wild fluctuations in rankings based on small sample sizes and provide a stable foundation for the model to build upon. This dynamic weighting ensures that the rankings are both informed by past expectations and responsive to current realities, offering a balanced and continuously self-correcting assessment of team strength.
Evaluating Prediction Accuracy and Calibration
The credibility of any predictive model hinges on its accuracy and calibration, concepts Pomeroy meticulously detailed on the podcast. Accuracy refers to how often the model correctly predicts the outcome of a game or a team’s future performance. KenPom consistently ranks among the most accurate predictors for college basketball, a testament to its robust statistical framework. However, Pomeroy also stressed the importance of "calibration." Calibration goes beyond mere accuracy; it assesses whether the model’s predicted probabilities align with the actual frequencies of outcomes. For example, if KenPom predicts a team has a 70% chance of winning a game, then over a large sample size, that team should indeed win approximately 70% of such games. A well-calibrated model is crucial for applications like sports betting and bracketology, where understanding the true likelihood of an event is as important as predicting the winner.
Pomeroy discussed the challenges of maintaining high calibration across an entire season, especially given the inherent unpredictability and variance in college basketball. Factors such as injuries, player development, and scheduling quirks can introduce noise into the data. Nevertheless, the continuous refinement of the KenPom algorithms, coupled with its transparent methodology, has allowed it to maintain a high degree of both accuracy and calibration, providing users with reliable probabilistic insights into the sport.
KenPom’s Enduring Influence on March Madness Bracketology
The conversation inevitably turned to March Madness, the NCAA men’s basketball tournament, where KenPom’s influence is perhaps most pronounced. For millions of fans filling out brackets, KenPom provides a data-driven alternative to gut feelings or media narratives. Its rankings offer a clear, objective measure of team strength that often highlights undervalued or overvalued teams compared to traditional seeding. Pomeroy’s insights help identify potential "upset" candidates by revealing teams whose underlying efficiency metrics are stronger than their seed suggests, or conversely, expose highly seeded teams that might be vulnerable due to statistical deficiencies.
The widespread adoption of KenPom metrics by casual and serious bracketologists alike underscores the growing sophistication of sports fandom. No longer content with just watching games, fans increasingly seek analytical tools to enhance their understanding and improve their chances in bracket competitions. Pomeroy’s work has democratized advanced analytics, making complex statistical concepts accessible and applicable for a broad audience during one of the most exciting events in sports.
Broader Analytical Trends Across the Sports Landscape
The Moneyball podcast, true to its name, extended its analytical lens beyond college basketball. Hosts Eric, Shane, and Adi discussed other compelling trends in sports analytics, demonstrating the pervasive impact of data science across various disciplines.
World Baseball Classic Odds and Analytical Scouting: The conversation touched upon the complexities of setting odds for international tournaments like the World Baseball Classic. Unlike domestic leagues with extensive historical data, these events feature rosters assembled from different professional leagues, making predictive modeling more challenging. Analytics play a crucial role in evaluating player performance against varied competition, projecting team chemistry, and assessing the impact of unique tournament rules. Data-driven scouting now informs roster construction and in-game strategy, moving beyond traditional scouting reports to incorporate advanced metrics on hitting, pitching, and fielding efficiency, even for international competitions.
NFL Preseason Workload Questions and Player Management: The hosts also considered how analytics are shaping preseason workload management in the NFL. With increasing emphasis on player safety and injury prevention, teams are leveraging data to optimize practice intensity, track player fatigue, and make informed decisions about rest and recovery. This data-driven approach aims to ensure players are in peak physical condition for the regular season while minimizing the risk of career-altering injuries. The discussion resonated with the broader theme of the Moneyball podcast, which in a separate episode on March 18, 2026, featured Brandt Tilis, EVP of the Carolina Panthers, discussing how analytics shape NFL team building, draft capital, quarterback contracts, and roster construction – all critical areas where data science now informs multi-million dollar decisions.
Tennis Dominance and Statistically Extraordinary Stars: The podcast further explored how advanced statistics illuminate the dominance of certain tennis stars. Beyond traditional metrics like aces and unforced errors, analytics can reveal patterns in serve placement, return depth, shot efficiency, and court coverage that distinguish top players. These insights help understand how players like Novak Djokovic, Rafael Nadal, or Serena Williams maintain their extraordinary performance levels, identifying specific statistical advantages that contribute to their sustained success. This granular data allows for more precise coaching adjustments and opponent scouting, pushing the boundaries of individual sports analysis.
The Wharton Perspective: Analytics as a Business Imperative
The choice of Ken Pomeroy as a guest on Knowledge@Wharton’s Moneyball podcast underscores the Wharton School’s commitment to exploring the intersection of data science, economics, and high-stakes decision-making, not just in traditional business but also in the rapidly evolving world of sports. The principles KenPom employs – efficiency analysis, predictive modeling, Bayesian updating, and calibration – are directly transferable to business strategy, financial forecasting, and risk management. The podcast serves as a platform to showcase how analytical rigor, when applied to complex systems like sports, can yield profound insights and competitive advantages. It reinforces the idea that understanding and harnessing data is no longer an optional skill but a core competency for leadership and innovation across all sectors.
The ongoing discussions on Moneyball, ranging from college basketball analytics to NFL team building and the financial rails of fintech (as seen in another March 18, 2026, episode featuring Wharton professor David Zaring on "The Fed’s Payment Rails and Fintech Access"), illustrate Wharton’s dedication to presenting cutting-edge research and practical applications of data-driven insights. Similarly, the exploration of AI’s potential to manage entire medical decision processes, featured in a related article on March 18, 2026, highlights the school’s broad engagement with disruptive technologies and their societal implications.
Conclusion: The Enduring Quest for Predictive Superiority
Ken Pomeroy’s appearance on Moneyball served as a timely reminder of the profound and continuing impact of advanced analytics on sports. His meticulous approach to college basketball, centered on tempo-free efficiency and the "four factors," has not only transformed how fans and professionals evaluate teams but has also set a benchmark for predictive modeling in sports. As March Madness approaches, the insights from KenPom.com will once again guide millions in their quest for the perfect bracket, demonstrating that in the modern era of sports, data and statistical acumen are as crucial to success as athletic prowess. The ongoing evolution of these models, coupled with the broader application of analytics across diverse sports, affirms that the quest for predictive superiority is a continuous journey, pushing the boundaries of understanding and decision-making in an increasingly data-rich world.
About the Author
