What do the San Diego Padres have in common with the fleet of garbage trucks dispatched each day by the city of Charlotte, N.C.?
The answer: Both have been the subject of mathematical formulas wielded by Lawrence Bodin, a University of Maryland at College Park business professor.
Bodin, a serious baseball fan and fantasy league player, set out last year to figure out how the Padres could best get through the expansion draft without losing important players. He has written a report on his research, entitled "Who's on First -- with Probability .4," to be published in the International Journal of Computer and Operations Research.
A caveat for Padres fans: The report doesn't name names due to disclosure rules by Major League Baseball. But Bodin knows, and is watching tonight's World Series game with more than passing interest.
The Columbia mathematician, 57, based his calculations on the "analytic hierarchy process," an approach he relies on in the classroom and as a consultant. The concept is simple, even though the precise methodology is more complicated. The goal is to arrive at sound decisions despite having to consider many independent factors.
Bodin uses this and other methodologies as a consultant specializing in vehicle scheduling. He helped Charlotte, for example, figure out how many garbage trucks it needs and how to deploy them. He's done similar work for the U.S. Postal Service, Federal Express Corp. and other clients.
Reconciling such disparate factors as vehicle reliability and employee absenteeism is not, at heart, so different from fielding a baseball team, he figured.
Traditionally, decisions about baseball players are based on batting averages, gut instinct and front-office intrigue. But Bodin thought he could improve on that mix, using a computer program first developed for the Pentagon in nuclear weapons planning.
The result, a suggested list of players to protect, intrigued at least one key figure in the front office of the team that would become this year's National League champion.
"It's a great tool," said Padres director of baseball operations Eddie Epstein. "If baseball people had more people like me as general managers, we would use it."
He said he collaborated with Bodin, an acquaintance, over a period of weeks last summer as an academic exercise.
Building a database
Epstein, a former director of research and statistics for the Orioles, supplied Bodin his evaluations of each player and what the Padres needed. Also factored: the team's desire for a rapid improvement, to help it win support for a new, publicly funded stadium.
Bodin plugged everything into his computer.
The resulting ranking suggested which players the Padres should protect from the Tampa Bay Devil Rays and Arizona Diamondbacks -- new clubs stocking their rosters for the first time in the Nov. 18, 1997, expansion draft. Each veteran club was allowed to shield 15 of its 42 draft-eligible players for the first round; a few more players could be protected after each round.
Bodin thought the situation was well-suited for the analytic hierarchy process.
The inventor, Thomas L. Saaty at the University of Pittsburgh, said yesterday that he came up with the system in the 1960s while working with government officials responsible for nuclear weapons and disarmament strategy.
The idea is to overcome a basic weakness in human cognition: how to weigh and compare many independent factors to choose a course of action.
In assessing the risks and values of various disarmament proposals, officials had to consider hundreds of factors, some subjective, others objective. He developed a system that breaks down a decision into its component factors and assigns priorities to each.
Although the system uses advanced matrix mathematics, it can be accomplished with a pencil and paper.
"It standardizes our thought processes the way a dictionary standardizes our language," Saaty said.
He's written nine books on the theory and co-developed software, called Expert Choice, which has been used by 20,000 business and government decision-makers. It was used by the Navy to pick scuba gear for its Seals, by Amoco to develop its long-term corporate strategy, and by General Motors Corp. to choose design alternatives.
Bodin thought the same system could be applied to baseball, a sport of nuance and judgment. Working with Epstein, he established four primary criteria for evaluating each player: age, desirability to the expansion teams, and potential to contribute to the Padres in the short term and long term.
The final two categories were divided into a number of sub-criteria, such as a player's value -- taking into account his salary -- and his leadership in the clubhouse.
Formula to reality
Epstein used his knowledge of each player to slot him within each criteria. The computer program assigned each player a numerical ranking. He and the Padres are not allowed to reveal whom they did or didn't protect. But of the 15 players shielded on the first round by the Padres, nearly all ranked high in Bodin's calculations, the professor said.
Two players were taken from the team: third baseman Gabe Alvarez, who was selected by the Diamondbacks in a pre-packaged trade to Detroit, and catcher John Flaherty, who went to Tampa Bay.
Neither player had a standout season. Flaherty's .207 batting average put him at the bottom on his team for the season. He had three home runs and four errors. Alvarez ended the year batting .231 with 19 errors.
Epstein said he found the experience useful, and would consider using the program in the future -- although he expects he would face resistance within baseball's tradition-bound culture.
"Baseball people are allergic to anything that is not arrived at by the visceral computer inside them," he said. "I'm open to any way to evaluate players."
Pub Date: 10/21/98