End Game

Kill a pursuit yourself and you will be credited. Drag out a pursuit to where it needs to be killed by someone else and you are punished. It's that simple. Responsibility for knowing when to stop lays with those engaged in the pursuit.

The burden is with those engaged in the pursuit. Original funding for an innovative pursuit would have allowed for many avenues of research, many options for funding, and great flexibility in how the pursuit is managed on a daily basis. It's very difficult for outsiders to know when the remaining avenues are just not worth pursuing. We could just be facing the last obstacle before the next great breakthrough.

Researchers on the ground are not so handicapped. They've faced many obstacles before and implicitly understand the potential of their remaining avenues of pursuit. They can step back and view the entirety of the pursuit and adjust their enthusiasm accordingly. If they're spending other people's money+, we need mechanisms that force them to objectively report the future chances of success. For example, we look to mechanisms for risk sharing, autonomous signals+ of their enthusiasm, independent on-the-ground evaluations, etc.

Further Reading