I first had the idea of representing risk versus payoff when trying to explain why DARPA (Defense Advanced Research Projects Agency, subsequently reverted to its original name "ARPA"), with 150 employees, had a different culture from that of the Defense Information Systems Agency, with thousands of employees and a different mission. I drew this figure for my presentation, "Exploiting DARPA C3 Initiatives," in The 1992 C3 Systems Engineering Conference (9-11 June 1992, DISA JIEO, Fairfax, VA).
In considering any project, you can make an assessment of the risk and payoff to locate the project. Experts may not exactly agree on coordinates, but their varying estimates should cluster close together. The object of making an investment in a project is, after doing some work, to move in the upper-left direction. If, in fact, your new knowledge drags you toward the lower right, you should forget sunk costs and scrap the idea. When payoff is very high compared to the risk involved, normal processes in the competitive sector are expected to see to the R&D without the need for government investment.
I have since expanded and generalized the illustration, towit:
© 1996 James
PottmyerWithin this scheme, what risks are considered acceptable by the people who must take them?
© 1996 James
PottmyerMy personal intuition is that there is a bimodal distribution of risk acceptability. Bimodal distributions are frequently encountered in evolutionary systems where there is advantage to diversity to accommodate large changes in the environment. We seem to have a mixture of two populations: a risk-averse one and a risk-taking one. Of many potential projects, relatively few can satisfy both risk-takers and the risk-averse. It is better to constitute teams of risk-takers separate from teams of risk-averse folks -- for the sake of the mental health of each group.