March Madness Metrics: What Makes a Champion?

By: Ella Hinkle and Taylor Fenton

Introduction

Every year, tens of millions of March Madness brackets are constructed to predict the outcome of the NCAA tournament. Out of the 68 teams, only one dominates the court. What makes these title-winning teams come out on top? Are all of their stats outperforming the average D1 team? Are they significantly better in certain defensive or offensive areas? By delving into the key traits of past winners, we can uncover the patterns and statistical factors that separate championship teams from the ones that fall behind.

Our analysis will focus on championship and runner-up teams' offensive and defensive metrics over the past 11 years. Specifically, we will categorize the following:

Offensive Strength
- Points per Game
- Shooting Efficiency
- Rebounds
Defensive Strength
- Blocks per game
- Defensive rebounds per game
- Two and three-point shooting percentage allowed
- Adjusted Defensive Efficiency

The teams analyzed will be the following:

Year	Championship Team	Runner Up
2024	UConn	Purdue
2023	UConn	SDSU
2022	Kansas	North Carolina
2021	Baylor	Gonzaga
2019	Virginia	Texas Tech
2018	Villanova	Michigan
2017	North Carolina	Gonzaga
2016	Villanova	North Carolina
2015	Duke	Wisconsin
2014	UConn	Kentucky
2013	Louisville	Michigan

Offensive Metrics

Points

The twenty teams competing for March Madness titles over the last decade have averaged 77.24 points per game throughout their seasons. Within the championship teams, the highest team average was 2021 Gonzaga, with 87.2 points per game, while the lowest was 2019 Virginia with just 57 points per game on average, as shown below. While there is an almost 30-point range between these teams in terms of points averaged per game, the majority of the other 18 teams’ averages were closer to ‘21 Gonzaga than ‘19 Virginia, with over 50% of the teams scoring above 78.1 points on average per game.

The largest grouping of these 20 championship teams averaged between 75 and 80 points per game, with 35% of teams falling within this category. While 30% of them averaged more than 80 points per game, there was still a significant 10% of teams that produced less than 70 points per game on average. With a standard deviation of 6.95 points, teams were pretty closely clustered around the average of 77.24 points. Over the past decade, the league average in points scored per game per team is usually about 70 points but has ranged from as low as 65 to as high as 78. This indicates that when compared to other D1 programs, the championship teams usually are above average, but the difference between averages in offensive scoring may not always be significant.

Shooting Efficiency

Shooting efficiency is evaluated based on a team’s field goal percentage throughout the season on average. On average, these 20 teams shot 47.88% from the field throughout their regular seasons, ranking them an average of 51st amongst D1 teams. This is a surprisingly low rank, but one that indicates that these teams likely also relied on strong rebounding efforts and/or strong defensive performances to carry them through the tournament. Additionally, the standard deviation of the team’s shooting efficiency rankings was 59.02, suggesting that there is a very large spread in terms of the team’s shooting performance relative to the rest of the league. Most of these teams cluster around 46-50% of attempts made, with that equating to about 26-30 baskets made each game.

2021 Gonzaga, depicted in the above graph as the point in the top right corner, scored on a whooping 54.9% of field goals attempted. As they were also the highest scoring of these teams on average, their offensive dominance certainly was critical in carrying them to the championship round. On the other hand, 2023 SDSU only converted 43.8% of their field goal attempts, a statistic ranking them 220th out of all D1 teams. It seems that for some teams, like 2021 Gonzaga, shooting efficiency was crucial to winning it all, but because of the large variation in data, this conclusion is certainly not true of every championship team.

Rebounds

Most of the 20 teams averaged between 24 and 28 defensive rebounds per game and averaged between 8 and 13 offensive rebounds per game. Averaging 15.75 rebounds per game, 2017 North Carolina, the data point on the far right on the below graph, averaged over 1.5 more offensive rebounds per game than any other championship-contending team over the last decade. Their offensive rebounding average ranked them #1 out of all teams during the 2016-17 season. This is significant because while their shooting efficiency was slightly below the average of these 20 teams at 46.6%, their average points per game was well above the average of these 20 teams at 84.4.

The largest percentage of teams ranked in the top 30 of Division 1 men’s basketball teams for offensive rebounds, making up 35% of these championship teams. However, almost an equal amount of these teams, 30%, were outside of the top 150 teams, putting them in the bottom half of the league overall. On average, these 20 teams ranked 103.4 overall, just squeezing into the top third of teams overall. So, for many teams, rebounding was a key driver of success, but for a similarly large amount, rebounding was not a crucial factor in winning it all.

Defensive Metrics

While offensive metrics seem somewhat inconclusive, we will dive into the defensive metrics of these top teams and how they compare with the rest of the tournament teams and runner-ups.

Blocks

Championship teams, within the past ten years, have averaged 4.27 blocks per game, while runner-ups have averaged 4.04. Six of the years, four being the past four, have seen the first-place team surpass the second in blocks per game. However, these teams' metrics are easily comparable as there isn’t a true difference between championship and runner-ups in blocks per game. Throughout the years, the difference in average blocks per game between first and second place finishers in March Madness have fluctuated with no solid difference.

Defensive Rebounds

Similarly, defensive rebounds per game – how many times the team gains possession of the ball after the offensive team misses a shot – is similar between championship and runner-up teams. Championship teams average 26 defensive rebounds per game, while second place averages 26.87. Interestingly, the boxplot shows that runner-up teams have a larger spread in defensive rebounds compared to championship teams, but this is most likely due to the small sample size of ten.

Shooting Percentage Allowed

Thus far, rebounds and blocks don’t illustrate any clarity that these title-winning teams are significantly different from their runner-ups. Since the final two teams are akin to each other in defensive rebounds and blocks per game – likely due to the fact that teams with weaker metrics have been weeded out by this point – we will look into how the championship team compares to the average D1 team. First, there is no instance from 2013 to 2022 where a champion has had both above the average two and three-point average shooting percentage allowed (of the 68 teams in the tournament). There are three years when the champion had above the average three points allowed, and four years above the average two points allowed. Interestingly, this may suggest that championship teams are worse in shooting percentage allowed.

Adjusted Defensive Efficiency

However, championship teams tend to have higher adjusted defensive efficiencies (ADJDE) than other teams in the March Madness tournament. Adjusted defensive efficiency measures how many points a team allows per 100 possessions. The x-axis on the graph below represents the round in which teams were eliminated from the postseason tournament; for example, S16 represents the teams that were eliminated in the sweet sixteen round of the March Madness tournament. As the round of elimination increases, the spread of the ADJDEs of the teams narrows and decreases. Championship teams average 90.8, surpassing the teams eliminated in earlier rounds, like teams in the round of 64 with an ADJDE of 98.24. Even compared to teams who get out in later rounds, like the Final Four with an average ADJDE of 92.04, they surpass them. This appears to be a more substantial metric to measure teams’ success in the NCAA tournament and can be a more comprehensive metric of tournament success.

Conclusion

From our analysis of the 20 teams who competed for NCAA Men’s basketball championships over the last decade, it is evident that some metrics have been more critical to their success than others. Some of these more important metrics include adjusted defensive efficiency, points per game, and field goal percentage made vs allowed. Rebounding and steals appear to be less significant.

Yet, at the same time, shortcomings in certain areas can be overcome by success in others. For example, some of these successful teams had lower shooting efficiency but made up for it by scoring a high amount of second-chance points.

Ultimately, if there was one metric, or even a combination, that could completely accurately predict which teams would contend for the championship, many perfect brackets would have been crafted over the years. March Madness highlights many of the things viewers love most about sports: high stakes, underdogs, and unexpected upsets. Even as our analysis can point to key metrics to consider when trying to pick a bracket winner for this current season, the only thing anyone can predict with absolute certainty is madness.