For someone who admits he is not an expert in the topics he studies and to "piggybacking" on predecessors' work, Bruce Bueno de Mesquita has become surprisingly influential.

A New York University professor, Mesquita has for decades been making predictions for the Central Intelligence Agency and the Department of Defense. He has even co-authored a book on political campaigns with former secretary of state Condoleezza Rice.

Today his predictive models are helping to reignite debate about the use of game theory beyond its normal habitat of mathematical and economic theory.

"It seemed to me, and others, quite natural to think about politics from game theory point of view," he told AFP, recalling his early steps into the field more than four decades ago.

Between 1982 and 1986 the CIA used his "Policon" model to make forecasts for over 30 countries.

In a now unclassified report, the agency said Mesquita's model "got the bull's-eye" twice as often as its traditional analysis. Taken together the two methods were accurate about 90 percent of the time, the agency said.

Two decades on Mesquita still has his sights on the world's hot button issues.

"My prediction on Iran is that they will not develop a nuclear weapon," he said with disarming confidence, "but they will develop enough weapons-grade fuel so that they can persuade people that they could build a bomb."

As he explains the relatively simple basis of his system -- listing actors and assigning each numerical values according to their policy stance and influence -- other predictions come thick and fast.

"I am quite optimistic that around the end of (President Barack) Obama's first term that there will be a serious peace agreement between the Israelis and the Palestinians, leading to the formal creation of Palestine.

"The contingency is that the current (Israeli) government, sometime in the next three years, will need to be replaced."

But he is less optimistic about the chances of a deal to reduce climate-changing carbon emission at Copenhagen in December.

Even if a deal can be reached, it will not work, he said, although the race to corner green technology markets may save the planet anyway.

The key to the model, he said, is understanding the self-interest of the parties concerned.

To convince North Korea to disassemble its nuclear weapons there is only one question, he said: "How much, quote 'aid' will be transferred to North Korea... to bribe the leadership into sticking to the agreement.

"The number that my model indicated was optimal was somewhere around one to 1.2 billion dollars a year."

Despite the obvious allure of predicting the future, some theorists urge caution.

Behind the apparent simplicity of Mesquita's work are the complex theories pioneered by John Nash, the Nobel-Prize-winning mathematician made famous in popular culture by the Oscar-winning film "A Beautiful Mind."

Nash's genius, according to his one-time colleague Avinash Dixit, was devising "a central organizing concept for games in which binding contracts cannot be reached."

In practical terms that means devising a status quo that is painful for actors to diverge from, despite the absence of a binding contract.

But Dixit, an economics professor at Princeton -- who has himself been tipped as a future Nobel Prize winner -- says some of the outstanding theoretical problems of game theory should limit practitioners' hubris.

"The theoretical analysis often assumed that every player was totally selfish, caring only for his or her own outcome," Dixit said.

"Increasingly we see in psychology and sociology that individuals have multiple goals, one of which might be concern for fairness."

He said practitioners should avoid black and white predictions, clearly marking out the range of possible outcomes in much the same way that meteorologists predict the path of hurricanes.

"Game theorists should be more humble or more careful and insist on producing these kind of cones of uncertainty around their forecasts," he said.

Another problem is the quality of data that goes into the model. In politics this means assigning quantities to abstract motives.

"The science tells you how to solve your model," said Dixit, "but what you put into your model is considerably an art."

For Mesquita the answer to this problem is simple: "If you put in garbage, you get out garbage."

"If you are looking at Israel-Palestine and your first cut of the data says Hamas is very eager for Israel to get whatever it wants, then there is something wrong with this data, you would know its garbage."

In the end, the only sure-fire way of testing game theory's usefulness may be to wait and see.

Share This Article With Planet Earth

del.icio.us	Digg	Reddit
YahooMyWeb	Google	Facebook

Related Links
Learn about nuclear weapons doctrine and defense at SpaceWar.com
Learn about missile defense at SpaceWar.com
All about missiles at SpaceWar.com
Learn about the Superpowers of the 21st Century at SpaceWar.com

Memory Foam Mattress Review

SPACE MEDIA NETWORK PROMOTIONS Solar Energy Solutions Tempur-Pedic Mattress Comparison ... Newsletters :: SpaceDaily :: SpaceWar :: TerraDaily :: Energy Daily XML Feeds :: Space News :: Earth News :: War News :: Solar Energy News