When the buzzer sounded and No. 11 seed Virginia Commonwealth University (VCU) defeated No. 1 seed University of Kansas to advance to the Final Four, it was a statistical anomaly.
Alexander G. Nikolaev, Ph.D., assistant professor in the department of industrial and systems engineering, has found that while it is unlikely for all four No. 1 seeds in the NCAA Men's Division I Basketball Tournament to advance to the Final Four, it is equally as rare for none to advance.
Each has occurred only once before, in 2008 and 2006, respectively.
This year's Final Four included zero No. 1 seeded teams as University of Connecticut (No. 3), University of Kentucky (No. 4), Butler University (No. 8), and VCU (No. 11) advanced.
Nikolaev's paper is based solely on statistical analysis of the pattern of seed advancement in the tournament from 1985 to 2010. The study's findings are not based on the characteristics of the involved teams.
"It doesn't matter how good a specific team is," Nikolaev said. "When they have to survive multiple stages [in the tournament], the probability of an upset becomes larger."
Although top-seeded teams typically win, the tendency for upsets throughout the tournament stems from the competition's game-to-game setup, according to Nikolaev.
"People love the Cinderella story of a [lower-ranked team] winning but that's not rare," said Mark H. Karwan, Ph.D., a professor in the department of industrial and systems engineering, and Nikolaev's faculty mentor. "What is rare is for the No. 4 or the No. 8 or the No. 11 to keep winning to get to the Final Four."
Nikolaev indicated that the probability of a No. 1 seed defeating a team ranked second is the same as the probability of a No. 2 beating a No. 3 or a No. 3 beating a No. 4, and so on.
The seed combinations most likely to advance are 1-1-1-2, 1-1-2-2, and 1-1-2-3, and it is most common for two No. 1 seeds to reach the finals, according to the analysis.
Nikolaev and co-researchers found that a geometric distribution, defined as the number of failures necessary before a success occurs, best represented the pattern of seed advancement.
"[The study] doesn't tell us what to do other than to expect at least one lower-ranked team to get to the Final Four," Karwan said.
Despite this, however, the professors reached a bracket-related consensus based on the data: it is wise to include a No. 1 team in a Final Four bracket.
However, Nikolaev pointed out that although data proves at least one No. 1 seed reaches the Final Four each year, it cannot predict which of the four No. 1 seeds will advance.
If fans want to ensure that they pick the correct No. 1 seed, they don't need to leave it to chance, they would need to pick all four, according to Karwan.
Students looking merely to make a successful bracket are still likely to select teams based on a subjective basis, rather than following statistical data.
"The only reason I'd ever pick all four No. 1 seeds to reach the Final Four is if I actually believed that they were the four best teams in the tournament," said David Fogel, a freshman communication major. "I pick my teams based on my gut feeling of who I think will perform well in the tournament."
A deciding factor for Fogel was player ability, a view that pure statistics does not allow for.
"I picked UConn because I believe they have the best college basketball player in the country in Kemba Walker," Fogel said.
It is difficult to separate opinion about how teams are performing from the actual likelihood that they will succeed in the tournament, according to Karwan.
Benjamin Veri, a freshman business administration major, also picked Connecticut to advance to the finals.
"There's always a possibility that all No. 1 seeds could make it," Veri said. "But with all the upsets that occur in college basketball, it's more likely [that] you're going to view the tournament in a more out-of-the-box way."
Over the weekend, Connecticut emerged as the champion, defeating Butler , 53-41.
Email: news@ubspectrum.com
Photo: Courtesy of Laron Fomby
