Column: The statistics behind NCAA Tournament success

NCAA Championship Trophy 2013

Last month’s 2013 NCAA Men’s Basketball Tournament final was one of the most exciting championship games in recent memory.

The contest featured the #1 overall seed Louisville Cardinals defeating the #4 seed University of Michigan Wolverines by a score of 82-76.  It was a back and forth affair that included many lead changes.

Louisville finished its season with an incredible surge, rattling off 16 straight victories. Along the way the Cardinals captured their third Big East Conference Tournament title, overcoming a 15-point deficit to defeat Syracuse in the final.

The Cardinals rode their streak to the NCAA crown. But does late-season success always translate into a strong performance in the NCAA Tournament? The Pleiad did some statistical analysis to answer that question.

We examined at the NCAA Tournament performance of the “power six” conference tournament champions. Given their late-season performance, those teams could be considered on a ‘hot streak’ entering the tourney. The study included results since 1998, when the Big Ten began it’s postseason competition.

In 11 of 16 years, or 68.7% of the time, a conference tournament champion also won the NCAA Tournament. This indicates a basic correlation between ‘hot streaks’ and NCAA Tournament success.

Of those 11 national champions, seven (64%) were #1 seeds, two (18%) were #2 seeds, and two (18%) were #3 seeds. Thus, only conference tournament champions that were high seeds in the 64 team NCAA Tournament have been able to capture the national title.

It is important to note that as high seeds, those championship squads were generally considered amongst the top 15 teams in the country entering March Madness. There were no Cinderella national champions.

Delving deeper, the Pleiad examined whether the performance of each conference tournament champion was over, under, or meeting expectations. We compared how far each team advanced in the NCAA Tournament to its projected placement, which is based on tournament seeding

For example, a #1 seed is projected make the Final Four since it is seeded as one of the top four teams in the tournament. Meanwhile, a #9 seed would not be expected to advance past the 1st round since it is seeded as one of the bottom 32 teams in the tournament.

The past 16 years have produced 92 “power six” conference tournament champions. Not included are four years when the Pac-12 did not hold a conference tournament. Overall, 20 (22%) of those teams met expectations, 32 (34%) exceeded expectations, and 40 (44%) performed under expectations.

Shockingly, since 1998 nearly half of all conference tournament champions have failed to meet expectations in the NCAA Tournament. This refutes the idea that ‘hot streaks’ are a good predictor of NCAA Tournament success.

But if winning your conference tournament doesn’t guarantee favorable results what does?

With the ‘hot streaks’ theory out the window, it’s useful to study the teams that won the NCAA Championship, despite losing their conference tournament.

In the past 16 years, the only teams to win the NCAA Tournament but not their conference tournament are Maryland in 2002, Syracuse in 2003, North Carolina in both 2005 and 2009, and Kentucky in 2012.

Other than Syracuse, each of those teams were #1 seeds in the NCAA Tournament. The Orange were a #3 seed. In fact, no team seeded lower than a #3 has won the national championship since 1998. What is the common denominator amongst all of these winners?

Talent.

Each NCAA champion was considered amongst the top 12 teams in their tournament based on seeding. Winning squads featured many future NBA stars like Richard Hamilton, Joakim Noah, and Carmelo Anthony.

Still, it’s worth noting that all 16 national champions enjoyed some success in their conference tournaments. Each team advanced to the semifinals, with the exception of Kentucky, who lost in the final.

But hold on a second. Michigan was mediocre at best at the end of the regular season, so how on earth did they make such a deep run?

The Wolverines limped to a 5-5 record in their last 10 games, including a  loss in the Big Ten tournament quarterfinals. Meanwhile, the biggest criticism of the Michigan all season had been their lack of experience.

Michigan’s playing rotation featured a number of freshmen logging big minutes. Nik Stauskas and Glenn Robinson III started for Michigan, while Mitch McGary, Caris LeVert and Spike Albrecht contributed off the bench.

All signs pointed towards an early NCAA Tournament exit. Yet inexperience and a late-season slump were not enough to overcome the Wolverines’ immense talent.

Michigan was led by Trey Burke, the consensus National Player of the Year. Burke topped the Wolverines in points (19.2) and assists (6.7) per game. The Wolverines’ lineup included three other players averaging double digit points as well.

Michigan also enjoyed a breakout performance from McGary, who averaged less than seven points per game during the regular season. The freshman scored 16.0 points per game in the NCAA Tournament, to go along with 11.6 rebounds.

Despite falling to Louisville in the national championship game, Michigan’s tournament performance only confirms the results of the Pleiad’s analysis.

Come March Madness, fans will salivate over teams that are entering the NCAA Tournament on a hot streak. Yet statistics indicate that conference tournament championships don’t automatically necessitate NCAA Tournament success.

It turns out that the best indicator of a potential national champion is simply the team’s level of talent.

Column co-authored by Geoffrey Knight

Photo courtesy of Adam Glanzman

About Dan Myckowiak 43 Articles
Dan is a senior Political Science major from Detroit, Michigan. He loves Detroit sports, and his favorite team is the Michigan State Spartans. Dan currently serves as editor of the Opinion section, and is formerly a managing editor, and editor of the Sports section for the Pleiad. Follow Dan on Twitter.

Be the first to comment

Leave a Reply

Your email address will not be published.


*