A U.S. organization is attempting to build a better prediction engine to help experts make more accurate decisions about world events.
Sponsored by IARPA, within the Office of the Director of National Intelligence, and supported by a team of academics, the project gives intelligence agencies an alternative to relying on experienced specialists.
The model, called Aggregative Contingent Estimation System (ACES), allows volunteers to compete to make the best predictions on topics such as politics, the military, economics, technology and world affairs. It will provide participants, who contribute anonymously, feedback on their successes. All questions are tied to an event that has a precise time frame, from weeks to months, so that the actual outcome can be compared to the forecast. In addition to participating in research, participants can use the tool to analyze their own forecasting skills, determining if they tend to be overconfident and which types of questions or topic-areas they are most successful in forecasting.
The system launched in beta format on July 15. According to the site, this system is different than other forecasting systems such as Intrade, in that it does not rely simply on averaging. Rather, scientists are using the site to test ways of combining group forecasts in a way that outperforms averaging. The site homepage encourages anyone to join, as long as they have interest in the topic, are English speaking, a college student or graduate and are over 18.
Questions range from “ estimate the change in spending on education in Sub-Saharan Africa” to “how many members will Facebook have by the end of 2011?” to “will the IOC select Annecy, France as the site for the 2018 Winter Olympics?” Perhaps the biggest question is what impact this and similar research sites will have. Will the results produce value or will it simply be a fun tool for participants? Will it create a trend, leading other industries to try to predict future business or medical events? Or, as one blogger proposes, will the project give validity to the old saying, “perception is reality?”