Wanna bet?
It was one of the worst ideas the Pentagon ever had. And it might have worked.
By DAVID BALLINGRUD, Times Staff Writer
© St. Petersburg Times
published September 14, 2003
It was an idea so evidently terrible that by the time it became public it was already well on its way to being scrapped.
On Monday, July 28, the news broke that Pentagon researchers were developing a sort of terrorism futures market - a way for Internet traders to bet on a host of military and political developments in the Middle East, including when the next terrorist attacks would come. A limited market was to open for business on Jan. 1.
As the bets rolled in, military planners said, they would gain important insights into future events - much in the way that a surge in stock purchases suggests good things are in store for a company traded on Wall Street.
The project was disclosed by its critics, of which there were many. Democratic senators assailed the terrorism futures market as "ridiculous," "grotesque" and "sick."
"I can't believe that anybody would seriously propose that we trade in death," said Senate Minority Leader Tom Daschle.
Embarrassed Pentagon officials couldn't pull the plug fast enough on the two-year-old program, which would have cost about $8-million.
But left in the dust of the futures-market fiasco were several unanswered questions. How exactly was the thing supposed to work? Would terrorists have been able to make a fortune wagering on their own actions?
Was it all as crazy as it sounded?
Some say it wasn't. Recently, a number of people who write about economics, technology and national security have come forward to say that the FutureMAP project got a bad rap; that, for a nation starkly confronted with the inadequacy of its conventional information gathering methods just two years ago, it may well have been worth a try.
* * *
Robin Hanson is an associate professor at George Mason University and an adviser to Net Exchange, the company that developed FutureMAP (Futures Markets Applied to Prediction) for the Pentagon. He still seems a little stunned by how fast the program collapsed.
"What was really striking is that on Monday, people had complaints, and on Tuesday we were cancelled," Hanson said in an interview last week. "Nobody even called us to clarify what the program was about."
Hanson explained how FutureMAP came to be and how it would have worked.
Desperate for information about terrorism after 9/11, and mindful of the fact that conventional spy work can never reveal everything, Pentagon officials turned to a flourishing Internet phenomenon called called prediction markets.
Prediction markets allow people to bet on a wide range of future events: oil prices, elections, wars, success of Hollywood movies and natural events such as storms or earthquakes. Individual traders may be in it for the money, but what interested the Pentagon is that the accumulated betting trends of such markets can give a fair picture of what is going to happen.
One, called the Iowa Electronic Markets, has had notable success predicting elections. Another, TradeSports.com, takes bets on everything from auto racing to rugby to the outcome of basketball star Kobe Bryant's legal troubles.
The Pentagon's program was put together by the Defense Advanced Research Project Agency, or DARPA, an agency charged with taking unconventional approaches to problems. DARPA worked with two private sector partners - Net Exchange and the Economist Intelligence Unit.
Though it became notorious as a prospective betting parlor for terrorism, and there would have been some of that, FutureMAP would have traded mostly in commonplace and measurable information - unemployment rates, currency values, tourism patterns - that taken together provide a snapshot of a nation's stability.
Specifically, traders would buy "futures contracts" on events they thought likely to happen within a given time frame, and sell off contracts as those events became less likely to happen.
For example, one might go online and see that the probability that Israel and the Palestinians would sign a peace treaty by the end of 2004 was set at 50 percent.
If the bettor thought that figure was low, he might wager that the probability was much higher, say, 90 percent, and purchase a futures contract reflecting that.
When the end of 2004 rolled around, and it was known whether or not there was a treaty, a bank under contract to the U.S. government would distribute money to the winners according to how confident they had been in their bets - the higher the percentage, the greater the payoff.
"Real money would have to be at stake," said Hanson, who worked on the project. "But not a lot of money. People would not be able to gain more than $100 per bet." The small payoffs would help discourage those who might try to play the market for profit.
The baseline percentages - the standards traders would be betting against - would have been established by the Economist Intelligence Unit, the business information arm of The Economist Group, publisher of the Economist.
There is some precedent for belief in the predictive power of markets, professor Robert Looney wrote this month in a publication of the U.S. Naval Postgraduate School's Center for Contemporary Conflict. He cited a study showing that, following the explosion of the space shuttle Challenger in 1986, the stock market immediately singled out manufacturer Morton Thiokol, well ahead of any public findings from the panel investigating the disaster.
But Looney cautioned that it's unclear how investors got their information, and he noted that the market was unable to repeat its fact-finding performance in response to the breakup of the shuttle Columbia in February.
* * *
Before it was killed, FutureMAP planned to begin registration of traders on Sept. 1. Yet according to Hanson, the question of who ultimately would have been allowed to play was never settled.
"Obviously we were trying for as much participation as we could," he said. "The concern was that we would open it up and nobody would come. So the plan was for a three-month trial for 100 or so invited people, playing with our money.
"On Jan. 1 we planned to open it up to the first 1,000 people, playing with their money. Then we would pause to see if the program could handle that many people." Eventually, he said, the program may have been open to anyone with a credit card and Internet access.
"The markets that have been the most dramatically successful have been opened to the widest possible participation," he said.
The question of whether terrorists might manipulate the market was never completely resolved either, Hanson said, but if the payoffs were kept small one incentive would be removed.
"How could we keep Osama bin Laden out of the program? I don't know that we could have," he said.
"There are two bad things that could happen. The first is that someone might give up information to make money. But if a guy tells us he is going to rob a bank, isn't that a good thing?"
The second, Hanson said, is that someone might be willing to lose money to obscure what the market would otherwise reveal. That could happen, he said, but research has shown that even with disinformation, a big enough market tends to return to its correct level anyway.
* * *
The short answer to what happened to FutureMAP is that the "terrorism futures" tag hurt. But DARPA has mostly its own Web site to blame for that.
FutureMAP's initial posting - now removed - referred to "special events securities," including the assassination of Yasser Arafat, the overthrow of the king of Jordan and a missile attack by North Korea.
"That's what got the fuss started," Hanson said.
In his article, entitled "DARPA's Policy Analysis Market for Intelligence: Outside the Box or Off the Wall," Looney agreed that the notion of betting on terrorism made the project a tough sell.
Most of the wagering "would have had nothing to do with terrorist attacks," he said. "Instead, markets were to be created in more mundane events like "How fast will Saudi Arabia's nonoil gross domestic product grow next year?'
"Or "What chance does Prince so-and-so have by the end of the year in ascending to the throne in Country X?' For these areas, there is no doubt considerable information out there that the United States does not have."
Hanson acknowledged, however, that the program might have included, on a case-by-case basis, wagering on "important and pivotal" global events, such as the signing of a peace treaty or an assassination.
The project's popularity was not helped by the fact it emerged from the DARPA office of John Poindexter, a central figure in the Iran-Contra scandal who last year was criticized for his involvement in a computerized surveillance project that would have gathered information about terrorist threats by probing financial, travel, medical and other databases.
Poindexter announced his resignation within days of the cancellation of FutureMAP, saying his work had been "distorted in press conferences and the media." He added, however, that some of the distortion occurred because of the use of "extremely bad examples" of the kind of information the program would produce.
Hanson regrets the examples as well. "The program was about a lot more than that," he said.
Perspective
Wanna bet?
Here golf courses grow like weeds
© 2006 • All Rights Reserved • St. Petersburg Times
490 First Avenue South • St. Petersburg, FL 33701 • 727-893-8111
|
|
