*********************************
There is now a CONTENT FREEZE for Mercury while we switch to a new platform. It began on Friday, March 10 at 6pm and will end on Wednesday, March 15 at noon. No new content can be created during this time, but all material in the system as of the beginning of the freeze will be migrated to the new platform, including users and groups. Functionally the new site is identical to the old one. webteam@gatech.edu
*********************************
Atlanta, GA | Posted: April 28, 2010
While Duke was crowned the victor in the NCAA men's basketball national championship this month, another big winner emerged: the Georgia Tech LRMC method for predicting NCAA tournament outcomes.
The new Bayesian LRMC method – an updated version of the previous system – correctly predicted the winner of more games during this year's tournament than all other ranking systems tracked on BCS computer ranker Ken Massey's website, masseyratings.com, which analyzes predictions for various sporting events.
Bayesian LRMC finished three full games ahead of the field, which included
well-known rankings such as the Associated Press and
USA Today polls, the NCAA's Ratings Percentage Index, Pomeroy,
Sagarin and Massey's own computer ranking methods. Bayesian LRMC was
the only ranking to correctly predict the winner of more than 50 games.
In addition to Bayesian LRMC's first place finish, the original LRMC method
finished in a three-way tie for second place with 48 correct predictions.
A correct prediction was defined as the winner of a game being ranked
higher than the game's loser.
This is the second recent success for the system. In 2008, the LRMC correctly
predicted the winner of every game in the final three rounds of the NCAA
tournament before the tournament started.
The LRMC method was first developed by Professors Paul Kvam and Joel Sokol at
Georgia Tech's H. Milton Stewart School of Industrial and Systems
Engineering (ISyE). The LRMC team has since expanded to include ISyE
professor George Nemhauser and City College of New York mathematics
professor Mark Brown.
A mathematical description of Bayesian LRMC is forthcoming in the Journal
of Quantitative Analysis in Sports.
