Game theory has been an important scientific tool to study the dynamics of social interactions n a community. Actually, my first publication is about the evolution of cooperation and how certain aspects of it can be modeled by the Iterated Prisoner's Dilemma.
I think it would be worthwhile to apply this tool to current developments in the science community. With funding being restricted (either overall or to certain topics) worldwide, competition between researchers also increases. This point was raised in a recent discussion over at ScienceBlogs.com on Open Notebook Science and fuels the debate on scientific misconduct.
Restricting funding increasingly to "hot" topics will lead to researchers aggregating in these topics like fish in a drying pond. Obviously, everybody will compete for the deepest spot. Moreover, the few researchers left in "cold" topics feel the pressure to "hype" their results in order to get funds. In the worst case, as a "cold" topic researcher is tempted to m(f)ake his topic "hot" with some surprising finding - not for fame or recognition, but to put food on the table. Few people are willing to fabricate data for fame, but if if you pressure them by essentially forcing them to decide between the highway and survival? Thus, with reference to open notebooks, how do you establish Open Science if there is currently an unprecedented incentive to fabricate data? How much less incentive does this current climate provide to share data than previous decades?
Indeed game theory can be applied exquisitely well to the research community. Currently all incentives are set to "cheat", yielding "all defect" as the only viable Evolutionary Stable Strategy. How do you switch off the pressure cooker, which only provides incentives to cheat and makes our results less reliable? Producing reliable data is difficult enough without fearing that you'll be out of a job if this particular results doesn't come out as a slam dunk on the front page of all major newspapers. Isn't it in the public interest that the research sponsored by tax money is at least reliable?