Dissecting KenPom: The Ultimate Guide to College Basketball Rankings

Understanding KenPom: A Deep Dive into College Basketball's Advanced Analytics

KenPom, short for Ken Pomeroy Ratings, has become a ubiquitous term in college basketball analysis. More than just a ranking, it's a comprehensive system that uses advanced statistical methods to evaluate team performance. This guide will provide an in-depth look at KenPom, breaking down its components, explaining how it works, and demonstrating how you can leverage it to improve your understanding of the sport.

What is KenPom and Why is it Important?

KenPom is a college basketball ranking system created by Ken Pomeroy, a meteorologist turned statistical analyst. Unlike traditional polls that often rely on subjective opinions, KenPom is purely data-driven. It focuses on measuring team efficiency, aiming to predict future performance based on past results. Its importance lies in its ability to provide a more objective and predictive assessment of team quality than traditional metrics.

• Objective Measurement: Eliminates biases present in human polls.
• Predictive Power: Designed to forecast future performance.
• Comprehensive Data: Considers virtually every Division I game.
• Informed Decision-Making: Aids in bracketology, game analysis, and understanding team strengths and weaknesses.

For example, a team might be ranked lower in the AP Poll due to a recent upset loss, but KenPom might still rate them highly due to their consistent efficiency throughout the season. This difference highlights the value of KenPom in providing a more balanced perspective.

Key Components of the KenPom Rating System

The KenPom rating system is built upon several key components, each contributing to the overall ranking. Understanding these components is crucial for interpreting the ratings effectively.

Adjusted Efficiency Margin (AdjEM)

The cornerstone of KenPom is Adjusted Efficiency Margin (AdjEM). This metric represents the difference between a team's Adjusted Offensive Efficiency (AdjO) and Adjusted Defensive Efficiency (AdjD). It's expressed as points per 100 possessions and adjusted for the strength of the opponent. A positive AdjEM indicates a team is expected to outscore its opponents, while a negative AdjEM suggests the opposite.

Formula: AdjEM = AdjO - AdjD

AdjEM is arguably the single most important number in the KenPom rankings. The team with the highest AdjEM is typically considered the best team in the country, and this rating is strongly correlated with success in the NCAA Tournament.

To illustrate, if Team A has an AdjO of 120 and an AdjD of 95, their AdjEM is 25. This means they are expected to outscore their opponents by 25 points per 100 possessions, adjusted for opponent strength.

Adjusted Offensive Efficiency (AdjO)

Adjusted Offensive Efficiency (AdjO) measures a team's points scored per 100 possessions, adjusted for the quality of the opposing defense. It provides a standardized way to compare offensive performance across different teams and schedules.

A higher AdjO indicates a more efficient offense. This metric is crucial for identifying teams that are consistently able to score points, regardless of the opponent.

For instance, a team with an AdjO of 115 is considered to be scoring 115 points per 100 possessions against an average Division I defense. This allows for a fair comparison even if the team plays in a conference known for its strong defenses.

Adjusted Defensive Efficiency (AdjD)

Adjusted Defensive Efficiency (AdjD) measures a team's points allowed per 100 possessions, adjusted for the quality of the opposing offense. It's the defensive counterpart to AdjO and indicates how effectively a team prevents its opponents from scoring.

A lower AdjD indicates a more efficient defense. This metric is essential for identifying teams that excel at stopping their opponents and forcing turnovers.

A team with an AdjD of 90 is allowing 90 points per 100 possessions against an average Division I offense. This is a strong defensive rating, suggesting the team is effective at limiting scoring opportunities.

Tempo

Tempo refers to the number of possessions a team averages per game. It's a crucial factor in interpreting offensive and defensive efficiency ratings. Teams with a faster tempo will have more possessions, potentially leading to higher raw point totals, while teams with a slower tempo will have fewer possessions.

KenPom uses adjusted tempo, which factors in the pace of the opposing team to provide a more accurate representation of a team's preferred playing style.

Understanding tempo is vital because a team with a high AdjO and a fast tempo might score a lot of points, but their AdjEM might be lower than a team with a slightly lower AdjO but a slower tempo and a better AdjD.

Luck

Luck, in the KenPom context, measures the deviation between a team's actual winning percentage and their expected winning percentage based on their efficiency margin. It aims to identify teams that have been unusually fortunate or unfortunate in close games.

Teams with a high luck rating have likely won more close games than expected, while teams with a low luck rating have likely lost more close games than expected. This metric often regresses towards the mean over time, suggesting that teams with extreme luck ratings are unlikely to maintain them throughout the entire season.

For example, if a team consistently wins close games despite being outplayed for portions of the game, their luck rating will be high. Conversely, a team that consistently loses close games due to unfortunate bounces or missed free throws will have a low luck rating.

Strength of Schedule (SOS)

Strength of Schedule (SOS) measures the difficulty of a team's schedule. It's calculated based on the KenPom ratings of the teams they have played. A higher SOS indicates a more challenging schedule, while a lower SOS indicates a less challenging schedule.

Understanding SOS is crucial for evaluating a team's performance. A team with a high AdjEM and a high SOS is generally considered more impressive than a team with a similar AdjEM but a lower SOS.

A team that plays a schedule filled with top-25 teams will have a much higher SOS than a team that primarily plays against lower-ranked opponents. This difference is factored into their overall KenPom rating.

Non-Conference Strength of Schedule (NCSOS)

Non-Conference Strength of Schedule (NCSOS) specifically measures the difficulty of a team's non-conference schedule. It's particularly important early in the season when conference play hasn't yet begun.

NCSOS can provide valuable insights into how well a team has prepared for conference play. A team that has faced a challenging non-conference schedule is often better equipped to handle the rigors of conference competition.

Teams that participate in prestigious early-season tournaments and schedule games against high-quality opponents will typically have a higher NCSOS.

How KenPom Ratings are Calculated: The Inner Workings

While the specifics of the KenPom algorithm are proprietary, the general principles are well-understood. The system relies on a regression-based model that analyzes game results and adjusts for various factors to arrive at the efficiency ratings.

Regression Analysis

The core of the KenPom system is regression analysis. This statistical technique is used to identify the relationships between different variables and predict outcomes. In the context of KenPom, regression analysis is used to predict a team's performance based on its past results and the performance of its opponents.

The regression model considers factors such as: points scored, points allowed, possessions, opponent strength, and game location. By analyzing these factors across a large sample of games, the model can estimate a team's underlying efficiency.

Iteration and Convergence

The KenPom ratings are not calculated in a single step. Instead, the system uses an iterative process that refines the ratings over time. Each game provides new data that is fed into the model, causing the ratings to adjust and converge towards a stable equilibrium.

This iterative process ensures that the ratings are constantly updated and reflect the most recent performance of each team. As the season progresses, the ratings become more accurate as more data becomes available.

Margin of Victory

While KenPom focuses on efficiency, it also incorporates margin of victory into its calculations. However, the impact of margin of victory is capped to prevent outliers from unduly influencing the ratings. This approach acknowledges that while winning and losing are important, the manner in which a team wins or loses can also provide valuable information.

The exact capping mechanism is not publicly disclosed, but it's generally understood that the impact of blowouts is diminished to prevent them from skewing a team's rating too dramatically.

Home Court Advantage

KenPom accounts for home court advantage by adjusting game results based on the location of the game. Teams typically perform better at home due to factors such as familiarity with the venue, crowd support, and travel fatigue for the visiting team.

The home court advantage adjustment is based on historical data and can vary slightly from season to season. This adjustment ensures that teams are not unfairly penalized for playing on the road or unduly rewarded for playing at home.

Using KenPom for College Basketball Analysis and Predictions

KenPom provides a powerful tool for analyzing college basketball and making predictions. By understanding the key components of the system and how they are calculated, you can gain a significant edge in your analysis.

Evaluating Team Performance

KenPom allows you to evaluate team performance more objectively than traditional polls. By focusing on efficiency metrics, you can identify teams that are consistently performing well, regardless of the opponent or the game situation.

When evaluating a team, consider their AdjEM, AdjO, and AdjD. A team with a high AdjEM and balanced offensive and defensive efficiency is typically a strong contender.

For example, a team ranked in the top 25 in both AdjO and AdjD is likely to be a well-rounded team capable of competing at a high level.

Identifying Overrated and Underrated Teams

KenPom can help you identify teams that are overrated or underrated by traditional polls. Teams with a high KenPom rating but a lower AP Poll ranking may be undervalued, while teams with a low KenPom rating but a higher AP Poll ranking may be overvalued.

This discrepancy can often be attributed to factors such as recency bias, conference affiliation, or media hype. By comparing KenPom ratings to traditional polls, you can gain a more nuanced understanding of team quality.

A team from a smaller conference that consistently outperforms expectations might be underrated by the AP Poll but highly regarded by KenPom.

Making Game Predictions

KenPom is designed to predict game outcomes. The system provides an expected score for each game, based on the efficiency ratings of the two teams involved and the location of the game.

To make game predictions using KenPom, simply subtract the away team's AdjEM from the home team's AdjEM and add the home court advantage adjustment (typically around 3 points). The result is the expected point spread for the game.

For instance, if Team A (playing at home) has an AdjEM of 20 and Team B has an AdjEM of 10, the expected point spread is 20 - 10 + 3 = 13 points. This suggests that Team A is favored to win by 13 points.

Bracketology and NCAA Tournament Analysis

KenPom is a valuable tool for bracketology and NCAA Tournament analysis. The system can help you identify teams that are likely to perform well in the tournament based on their efficiency ratings and other key metrics.

Historically, teams with high KenPom ratings have had a higher success rate in the NCAA Tournament. Teams ranked in the top 25 in KenPom are often considered strong contenders to make a deep run.

When filling out your bracket, consider the KenPom ratings of each team, as well as their strength of schedule, luck rating, and other relevant factors.

Identifying Potential Upsets

While KenPom is generally accurate in predicting game outcomes, it's not infallible. By identifying teams that are likely to outperform their KenPom rating, you can potentially identify potential upsets.

Teams with a low luck rating or a strong non-conference strength of schedule may be poised for a breakthrough performance. Additionally, teams with a significant coaching advantage or a unique playing style may be able to overcome their KenPom rating.

Consider factors such as injuries, suspensions, and team chemistry when assessing the potential for upsets. These factors are not directly captured by KenPom but can significantly impact game outcomes.

Limitations of KenPom and Considerations for Use

While KenPom is a powerful tool, it's important to recognize its limitations. The system is based on statistical analysis and does not account for all the factors that can influence game outcomes.

Data Dependence

KenPom relies on historical data to generate its ratings. This means that the system is most accurate later in the season when more data is available. Early in the season, the ratings may be less reliable due to the limited sample size.

Be cautious when using KenPom to make predictions early in the season. As more games are played, the ratings will become more stable and accurate.

Lack of Qualitative Factors

KenPom does not account for qualitative factors such as team chemistry, coaching ability, player motivation, or injuries. These factors can significantly impact game outcomes but are not directly captured by statistical analysis.

Consider these qualitative factors when making predictions, especially in high-stakes games where emotions and motivation can play a significant role.

Potential for Regression to the Mean

Metrics like Luck can be volatile. Teams with extremely high or low luck ratings are likely to regress towards the mean over time. Be wary of relying too heavily on these metrics when making predictions.

Focus on the underlying efficiency ratings (AdjEM, AdjO, AdjD) as these are more stable and predictive over the long run.

Impact of Rule Changes

Significant rule changes can impact the efficiency ratings and predictive power of KenPom. For example, changes to the shot clock or the free throw rules can alter the pace of play and scoring patterns.

Be aware of any rule changes that may affect the game and adjust your analysis accordingly. KenPom typically adjusts its algorithms to account for these changes, but it's important to be aware of the potential impact.

Tournament Variance

The NCAA Tournament is inherently volatile, and upsets are common. While KenPom can help you identify teams that are likely to perform well, it cannot predict every outcome with certainty. Single-elimination tournaments are prone to variance, and even the best teams can be vulnerable to an off night.

Don't put too much stock in any single prediction. Embrace the unpredictability of the tournament and enjoy the ride.

Advanced KenPom Strategies and Considerations

Beyond the basics, there are more advanced strategies to using KenPom to refine your analysis.

Trend Analysis

Analyzing trends in KenPom ratings can provide valuable insights. For example, a team that is consistently improving its AdjEM over time may be poised for a strong finish to the season.

Track KenPom ratings over time to identify teams that are trending upwards or downwards. This can help you anticipate future performance and identify potential sleepers.

Conference-Specific Analysis

KenPom can be used to analyze the relative strength of different conferences. By comparing the average KenPom ratings of teams within each conference, you can assess which conferences are the most competitive.

This information can be useful for evaluating the performance of teams from different conferences and making predictions about inter-conference matchups.

Combining KenPom with Other Metrics

KenPom is most effective when used in conjunction with other metrics and qualitative analysis. Don't rely solely on KenPom when making predictions. Consider factors such as injuries, coaching, and team chemistry.

Other valuable metrics include: NET rankings, Sagarin ratings, and statistical data on individual player performance.

Understanding Bayesian Priors

KenPom, especially early in the season, utilizes Bayesian priors. This means it starts with a prior expectation of team performance and updates it with new data. This helps stabilize the rankings with limited data.

Be aware that early-season rankings are influenced by these priors, which are based on historical data and pre-season expectations. As the season progresses, the impact of the priors diminishes.

Real-World Examples of KenPom in Action

To solidify your understanding, let's look at some real-world examples of how KenPom can be applied.

Example 1: Identifying a Tournament Sleeper

In the 2018 NCAA Tournament, Loyola Chicago emerged as a Cinderella story, reaching the Final Four as an 11-seed. While many were surprised by their success, KenPom suggested they were a potentially dangerous team. They had a strong AdjEM, indicating a solid underlying efficiency, and a relatively low luck rating, suggesting they were due for some positive regression. This combination made them a prime candidate to outperform their seed.

Example 2: Spotting an Overrated Team

In a previous season, a high-major team received significant media attention and a high AP Poll ranking despite having a mediocre KenPom rating. Their success was largely attributed to a few high-profile wins, but their underlying efficiency metrics were not impressive. As the season progressed, their performance declined, and they ultimately underperformed expectations, validating the concerns raised by KenPom.

Example 3: Predicting a Conference Champion

In a particular year, the Big Ten conference was considered one of the strongest in the country. KenPom consistently ranked several Big Ten teams among the top 25 nationally. By analyzing the KenPom ratings of the top teams in the conference, it was possible to accurately predict the eventual conference champion and identify the teams that were most likely to receive high seeds in the NCAA Tournament.

Conclusion: Mastering KenPom for College Basketball Success

KenPom is a powerful tool for analyzing college basketball and making predictions. By understanding the key components of the system, how they are calculated, and its limitations, you can gain a significant edge in your analysis. Whether you're a casual fan, a serious bettor, or a college basketball analyst, mastering KenPom can help you make more informed decisions and deepen your appreciation for the sport.

Remember to use KenPom as part of a broader analytical approach, combining it with other metrics, qualitative factors, and your own knowledge of the game. With practice and experience, you can unlock the full potential of KenPom and become a more informed and successful college basketball observer.