By Danny Blumenthal
There is no one path to the championship in college basketball. Some teams shoot the lights out on offense, others ride a suffocating press to success, and still others are decent on both offense and defense. Nevertheless, the mantra that “Defense Wins Championships” has never gone away. Many pundits believe that defense is more consistent than offense, and that a solid defense can bail out a team when its shots aren’t falling – a crucial asset in an elimination tournament. On the flip side, numerous changes to the game, such as the spike in three-pointers, seem to have given offenses more influence than defenses. To identify the best strategy for March Madness success, it’s time to explore the different types of teams.
Anecdotally, teams can win it all in a variety of ways. While Virginia has been famous for its defense in recent years, it was quite balanced in 2019, ranking 2nd in KenPom’s adjusted defensive efficiency and 5th in adjusted offensive efficiency. And in the years prior to that, 2018 Villanova (22nd on defense, 1st on offense), 2017 UNC (25th on defense, 4th on offense), and 2015 Duke (37th on defense, 3rd on offense) all demonstrated that even if teams are only “pretty good” on defense, they can win a title on the back of an elite offense. Finally, UCONN showed the opposite trend in 2014. Although they entered the tournament ranked 57th on offense, they rode a stingy defense (12th in the country) all the way to the title. With all of these potential paths to the title, it’s fair to ask whether one strategy is better than the rest. Do teams with elite offenses or defenses succeed more? Is it better to dominate one aspect of the game, or should teams aim to be merely above average on both ends of the floor?
To answer this question, data on each team’s adjusted KenPom efficiency margins since 2002, as well as their Performance over Seed Expectation (PASE) and tournament success were collected. PASE controls for a team’s expected performance and identifies whether it exceeded or fell short of previous teams with that same seed. For example, 5th-seeded Auburn would have been expected to win about 1.1 games in the 2019 tournament. Instead, they advanced to the Final Four by winning four games, thus earning a PASE of +2.9. Teams were then split into two categories: those with an absolute gap of 50 or more places between their adjusted offensive efficiency rank and adjusted defensive efficiency rank (“Unbalanced”) and those with a smaller gap between offense and defense (“Balanced”). Importantly, the balanced teams are not just teams with elite offenses and elite defenses. They also include lower-ranked teams, such as this year’s Norfolk State squad, which ranked 204th in the country on offense and 219th on defense.
Finally, a model predicting PASE from team balance, controlling for factors such as overall adjusted efficiency margin and seed, was implemented. The results (with balanced teams used as the reference group) are displayed below:
Teams with a higher adjusted efficiency margin tend to exceed their seed’s expectations, since they have likely been under-seeded by the selection committee. Conversely, teams with worse seeds tend to have higher PASE’s, but this is likely due to a floor effect. Weaker teams can’t underperform their expectations by much, while there is a long way to fall should a highly-ranked team get upset early. Still, even after controlling for these factors, unbalanced teams underperform significantly more than balanced teams. This suggests that balanced teams will play better in March than those which excel in only one area of the game.
Taking this one step further, are there differences between teams with a focus on offense or defense? The same model was run again, but this time, teams were split into three categories: “Offensive Focus”, “Defensive Focus”, and “Balanced.” In this case, while both offense-oriented (0.15 fewer wins over expected) and defense-oriented (0.12 fewer wins over expected) teams were worse than balanced teams, only offense-oriented teams significantly underperformed more often than balanced teams. The following graph displays the differences between the three types of teams.
If unbalanced teams, especially those with much stronger offenses than defenses, are more likely to fall flat in the tournament, fans should be aware of several teams before they fill out their brackets. The table below indicates the top unbalanced teams, as well as whether they focus more on offense or defense.
Teams such as Kansas and Tennessee could be on upset alert soon. They both boast top-ten defenses, but sit outside the top 50 on offense. Unless they can get their offenses rolling, they may be in for a shorter-than-expected stay in Indianapolis. Meanwhile, Ohio State faces the opposite problem. While the team ranks 4th in adjusted offensive efficiency, they are only 79th on defense, and no team in the KenPom era has ever made the national title game with as bad a defense as the Buckeyes. And while their gap is just short of 50 spots, Iowa too is overly-reliant on its offense. The Hawkeyes boast the 2nd-best offense in the country, but teams like Indiana, Illinois, and even lowly Minnesota were able to beat the 50th-ranked defense by lighting up the scoreboard.
On the other end of the spectrum, a few teams stand out as particularly balanced and underrated by the committee. UCONN was only seeded 7th in the East region, but they are one of only six teams in the country to rank in the top 25 in both adjusted offensive and defensive efficiency. BYU was only given a 6-seed, but they too possess great balance. Ranking 28th on offense and 26th on defense, the Cougars could give a top team a scare, just as they did to Gonzaga in the WCC title game. These teams can succeed on both ends of the floor, and no matter what their opponent throws at them, they will be prepared.
Overall, it appears that balanced teams perform better in March than unbalanced teams, especially those which rely more heavily on offense. While the difference between balanced and unbalanced teams is not major, it’s enough to move a team from an underdog to a slight favorite. In particular, this information could be valuable when two teams seem comparable on other metrics – such as 4-seed Purdue (23rd on offense, 23rd on defense, 13th in overall efficiency) and 5-seed Villanova (9th on offense, 68th on defense, 12th overall). NBA fans would do well to consider balance too, since teams such as the Brooklyn Nets (1st on offense, 25th on defense) may not be able to sustain their success in the playoffs. If you’re unsure of which team to pick for your bracket this week, remember what the motivational posters from school said: balance is the key to success.