MACON — Allen Lynch, Associate Professor of Economics at Mercer University, along with Jay Coleman, Professor of Management at the University of North Florida have been predicting recipients of at-large bids handed out by the NCAA Tournament Selection Committee for the more than a decade. The “Dance Card,” now in its 13th year, has a 93.7 percent accuracy in predicting the “at-large” berths handed out by the NCAA Tournament Selection Committee.
In the last two years, however, they have added a model to predict the results of individual tournament games once the madness begins. Lynch and Coleman’s new model, dubbed “Score Card,” predicts the winners of tournament games. Although predicting the winner of individual games represents a much more daunting task than predicting tournament entry, Lynch and Coleman still forecast winners reasonably well. Information about each model appears below.
The professors used the results of the 256 tournament games played between 2001 and 2004 to devise the Score Card formula as an estimate of which pieces of information were most important in determining who wins games in the tournament, and the weights that are placed on those pieces of information. After developing the Score Card model the professors’ applied the model to the 2001-2004 tournaments and found that it was
75 percent accuracy overall, 78 percent accurate in the first round, 72 percent accurate in rounds 2-6 and 66 percent accurate in “close” games (i.e., games in which the two teams are within 10 spots of each other in the RPI ranking).
And the model also was accurate at predicting upsets: 74 percent accuracy whenever it predicts an upset (i.e., when it predicts that the team with the higher RPI has less than a 50 percent chance of winning the game) and 86 percent accuracy whenever it predicts a more certain upset (i.e., when it predicts that the higher-ranked team has less than a 40 percent chance of winning the game).
For more information on the Score Card Formula, visit the professors’ web site: www.unf.edu/~jcoleman/score.htm
The Dance Card is a formula derived by Coleman and Lynch as an estimate of the decision rule that the NCAA Tournament Selection Committee uses when picking the teams that will get at-large bids to the NCAA men’s basketball tournament. Over the past 13 years, the Dance Card formula has correctly predicted 416 of the 444 available at-large Tournament slots, or 93.7 percent. Over the 7-year period, 2000 through 2006, for which the Dance Card has been used since its initial development in 1999, it has correctly predicted 224 out of the 239 available at-large Tournament slots.
Lynch and Coleman, the Richard deRaismes Kip Professor of Operations Management and Quantitative Methods at UNF’s Coggin College of Business, published the Dance Card formula in Interfaces, a journal of the Institute for Operations Research and the Management Sciences, in 2005.
Coleman and Lynch will be making this year’s final predictions on Selection Sunday, but their current analysis, which is updated regularly, can be found at Dancecard.unf.edu.
About Mercer University:
Founded in 1833, Mercer University is a dynamic and comprehensive center of undergraduate, graduate and professional education. The University has 7,300 students; 11 schools and colleges – liberal arts, law, pharmacy, medicine, business, engineering, education, theology, music, nursing and continuing and professional studies; major campuses in Macon and Atlanta; four regional academic centers across the state; a university press; two teaching hospitals — Memorial Health University Medical Center and the Medical Center of Central Georgia; educational partnerships with Warner Robins Air Logistics Center in Warner Robins and Piedmont Healthcare in Atlanta; an engineering research center in Warner Robins; a performing arts center in Macon; and a NCAA Division I athletic program. For more information, visit www.mercer.edu.
Editor’s Note: Media interested in interviewing Lynch and Coleman or having them on air should contact Mark Vanderhoek at (478) 301-4037 or (800) 837-2911. Both can be available on “Selection Sunday.”