A sortition-like system for grants would help for sure. The problem is, how many proposals are you permitted to submit? If you flood the process you stack the d

A sortition-like system for grants would help for sure. The problem is, how many proposals are you permitted to submit? If you flood the process you stack the deck for a random win in your favour. I think random reviews from professionals outside of a field might be possible as well, at least reviewing to what extent the data supports a hypothesis. For instance, say you had a random set of statisticians review the datasets and accompanying hypotheses for 2 or 3 competing theories in biology and cast their votes on what hypothesis seems most plausible given their assessment of the data. They wouldn't be familiar enough with the field to recognize the "leaders" that might be influencing a field politically rather than empirically. Not a perfect system for sure, so maybe someone has a better idea.