Joshua Nicholson and John Ioannidis suggest that the US National Institutes of Health (NIH) is failing to fund the most impactful researchers in favour of more conformist scientists with interests similar to those of the grant reviewers (Nature 492, 34–36; 2012). For presumably practical reasons, Nicholson and Ioannidis define these high-impact scientists as authors of a paper(s) with at least 1,000 citations. But setting the bar so high could yield unrepresentative outliers in the analysis.

Many of these authors are unlikely to repeat this level of impact consistently, so their papers become anecdotal rarities and atypical of career performance. And as the Comment revealed, several authors of papers with 1,000 or more citations had varied and justifiable explanations for not holding NIH funding. Also, agencies such as the Howard Hughes Medical Institute skim off some of the high-performing researchers, effectively removing them from the NIH system because they are considered to be well funded.

There can be network or familiarity bias among grant reviewers towards certain applicants, often for legitimate reasons — such as when an applicant's proposal might benefit from wider collaboration. Indeed, a junior investigator could find it productive to participate in a grant-panel study section, in which he or she can be exposed to new ideas and learn from peers.