I’ve been seeing in professors the result of the cutoff and scoring obsession and weird focuses of the NIH grant system. If I were head of NIH, I would say “We’re obsessing about the scores of these grants to level that exceeds our ability to tell good research from bad. How bout we just take an amount of grants that’s 4x the amount we can fund, and just randomly draw ¼ of those.“
Also, when I am Comrade General Secretary of the Socialist States of America, that’s how I’m allocating capital.
> not assigning members of your government allocation funding blocks that they can bet on research outcomes
Bad post OP
What would they bet on?
I’m joking and responded without actually thinking about it, but you’d probably use a broad basket of metrics that correlate with research success/value, which are semi-randomly weighted in ways that aren’t fully revealed to the “investors” in order to wreck min-maxing and attempts to short-circuit the metrics.
Grants would be composed of funding allocation blocks from multiple individuals, and those who did well on previous research funding would obtain more of them over time.
It’s similar to a market, only if you ripped out the utility function of the market and replaced it with something else.