By now, you have heard the narrative for nearly three weeks: Syracuse did not belong in the NCAA Tournament. With each unexpected Orange victory, the righteous indignation has only grown louder. ESPN’s Mike Greenberg even went as far as to say that Syracuse winning the National Championship would make the season a “joke.”
Let’s put aside the fact that the majority of “professional bracketologists” had Syracuse in the Tournament on Selection Sunday and that there were at least two teams with equally, if not more, questionable resumes. In fact, let’s put aside the question of traditional “Tournament resumes” entirely.
According to the NCAA, at-large spots in the Tournament are to be given to “the 36 best teams.” The Committee does its best job of determining this using the RPI, season-long results, and numerous other evaluations including simply “watching games.”
But are the top 36 (or 68) teams in any metric or combination of metrics at the end of the regular season truly the best? Or are those teams simply looking the best at that moment based on a sample of about 30 games?
If you are to listen to someone like Greenberg, it sounds like the games in the Tournament don’t count toward determining how good a team is. The sacrosanct regular season has already shown us who the good teams are, while March Madness is simply a wildly unpredictable (and entertaining) way of trying to validate it.
To an extent, he’s right: 6 games is too small a sample size to determine the objectively best team. But the college basketball regular season is also too small a sample size to determine the best team.
Taking Syracuse as an example, the Orange’s KenPom ranking bounced between 27 and 70 throughout the regular season. SU finished No. 41 after losing five of its last six games, dropping the team 10 spots. Was this arbitrary stopping point after 32 games the best time to determine Syracuse’s true worth?
Probably not, but you have to stop somewhere. Major League Baseball stops after a grueling 162-game regular season. According to a study by Los Alamos National Laboratory, however, it would take 256 games to ensure the best MLB team had the best record.
Most sports do not have enough time to truly determine the best team, so they create imperfect playoff systems instead (and also because they’re more entertaining). A study by Harvard University showed that of the four major professional sports, MLB’s playoff system is the worst (essentially random), while the NBA’s does the best job of affirming regular season results.
The NCAA Tournament, although not analyzed in the study, would likely fall closer to the MLB end of the spectrum, but there are two key points that commentators like Greenberg miss:
- While playoffs are by definition a small sample size, they also add to the overall sample of a season.
- A team’s skill level or the abstract concept of “best” is a moving target.
Pertaining to #1, Syracuse is now up to No. 22 in the KenPom rankings after its four-game Tournament run. This is easily the lowest rank of any of the Final Four teams, but it’s only two spots below Duke. Would Duke winning the National Championship have made the season a joke?
Pertaining to #2, a common criticism of Syracuse over the last week is that if they weren’t lucky at the very least they “got hot at the right time.” Numerous studies have shown that getting “hot,” whether in a single game or over a stretch of games, is simply something our minds make up to explain a random streak. A player is no less likely to miss a shot when he’s “hot”—it just seems that way.
Syracuse’s performance in the Tournament thus far might indicate they have improved as a team from the regular season. And while this may have come at a fortunate time, it does not make it any less meaningful. The current four-game winning streak contributes just as much to what we know about how good the team is in the abstract as its four-game winning streak to start the season or its losing three of the last four.
If Syracuse is able to win its final two games and take home a National Championship, it will almost certainly move into the top 15 in the KenPom rankings and possibly higher. Will this be any more of a “joke” than Duke winning last year from the No. 4 position? Connecticut from No. 8 and No. 9 in the last five years?