DARPA Network Challenge Proves that Prizes Rule

I heard about the challenge posed by the Darpa Network Challenge on a radio programme on the BBC and was intent on checking it out earlier this week. In a nutshell, Defence Advanced Research Projects Agency put of a competition with a prize of US$ 40,000 and a nine day time limit during which teams were expected to identify the precise locations of ten weather balloons randomly placed in public places within the United States.

My interest in this was peaked especially because of an ongoing interest I market design and the use of prizes as an incentive to developing solutions to business or public affairs. While I was sure that the prize would eventually be claimed, I am amazed at how quickly the contest ended. A team of MIT scholars claimed the prize within nine hours of its start and that is really efficient. As this story in the Guardian states, the team utilized incentives by splitting up portions of the prize and promising a windfall of US$ 2000 to anyone who send a verified clue of the location of a balloon. And to ensure that even those who do not directly identify a balloon are motivated to send the information around, there were smaller prizes for guiding the ultimate identifiers to the MIT team's website.

The full results about the organization of the search are not out yet but it definitely has other applications for it is a very clever design. What I find curious is that the team was sufficiently confident that it could share up to 50% of the total prize with the identifiers of the balloon and sit and wait for the correct and accurate information to reach it. making many assumptions, I am thinking that the MIT team could still have applied a reverse auction by inviting identifiers and asking them to auction their information. This would have had interesting results because it is possible that the price for identifiers could possibly have been less than US$ 2000.