Defense Agencies Set Budget Priorities ‘Moneyball’ Style
If professional sports teams can use computer-generated analysis to pick the best ball players, why couldn’t government bureaucracies apply the technology to make wiser investments?
In fact, Moneyball-style decision making is taking place across the U.S. government, most notably at military organizations whose budgets are being squeezed and that find themselves under growing pressure to deliver more bang for the buck.
Known as “collaborative decision making” software, this technology — which rose to fame thanks to Michael Lewis' best-seller "Moneyball: The Art of Winning an Unfair Game" — crunches data but does not replace human brains. The software aids managers in making better choices, said John Saaty, CEO of Decision Lens, a CDM software provider in Arlington, Va.
Decision Lens recently was chosen by the science-and-technology branch of the Marine Corps’ program executive office for land systems to allocate resources across a $9 billion portfolio that includes many of the Corps’ major weapon systems. Other clients include the Joint Staff, the Department of the Navy, the Army Research Development and Engineering Command and the Joint Improvised Explosive Device Defeat Organization.
With military budgets for new weapons projected to shrink, managers are hoping that CDM software can help them invest money so that it results in more useful equipment for troops in the field. A Marine Corps program manager who worked with Decision Lens told National Defense that he believes this technology could lead to smarter spending and could preclude agencies from wasting money on "shiny objects" promised by contractors.
“Typically it’s the person who bangs the table the loudest at the meeting that usually gets their way, or depending on how many stars they have, or a vendor demonstrating the latest whiz-bang shiny object. Next thing you know, we are funding this,” said the program manager, who was speaking on the condition of anonymity.
The Decision Lens approach, he said, removes some of the subjectivity from the process. “Based on agreed-upon criteria by the community, we can evaluate [candidate technologies] on a level playing field,” he said. “I expect to see a higher return on investment and a higher probability of transitioning to the war fighter.”
Decision Lens software is used by major corporations such as Kraft, Johnson & Johnson, Genentech and Siemens. It is also employed by professional sports clubs for player selection, including the Green Bay Packers, Oakland A’s, Arizona Diamondbacks. Calgary Flames and North Carolina Hurricanes.
In an interview, Saaty cautioned that the government’s use of CDM is far different than how the Oakland Athletics selected their players, as documented in Lewis' book and movie "Moneyball.”
Teams thought there was magic in data that was going to tell them who the best players were going to be based on statistical analysis, Saaty said. But that was not the case. “Decision Lens brings together the quantitative elements combined with the subjective judgments of the scouts.”
“Every decision is subjective,” said Saaty. “Even data has to be interpreted. … You’re not going to get all the answers from a black box.”
Decision Lens charges anywhere between $80,000 and $900,000 for its web-based collaborative software license, depending on the size of the organization. Army RDECOM, for instance, paid $700,000.
Skeptics might dismiss CDM software as a gimmick that offers no guarantees of success. Saaty said the return on investment “depends on the leadership of the organization.”
Some managers, he said, are “resistant because they have an autocratic style of making decisions.” But increasingly he sees more acceptance of this product as a means to put analytical rigor into judgment calls that typically have been made by people sitting around a table and advocating for their programs using spreadsheets and PowerPoint charts. “That’s not good enough anymore,” said Saaty. “You have to have better analysis.”
Technology obviously can’t be regarded as a panacea, he warned. “The organization has to invest intellectual capital in making this a success.”
Decision Lens’ software was born in the 1970s as a mathematical model that was designed by Saaty’s father — a mathematician who created the theory known as “analytic hierarchy process.” Saaty Sr. at the time worked for the State Department and led an arms-control team that was tasked with negotiating nuclear arms reductions with the Soviet Union. “My father realized his team had no framework to evaluate priorities,” said Saaty. “He was humbled by the experience … and developed a theory of prioritization.” Saaty and his brother later turned their father’s invention into a business by codifying the theory into a software platform.