On rejection rates

What a 12% acceptance rate has taught me about doing science.

I keep a public failure chronicle on the homepage. The current count is 2 acceptances out of 17 submissions — 12%.

The point of writing it down isn’t to perform humility. It’s that I genuinely cannot remember, in the moment, whether the work is good. The chronicle is a memory aid: this has happened before, you have been wrong about what reviewers will say before, the field has changed its mind before.

Three things I’ve stopped doing since I started keeping the count:

1. Reading rejections the same day they arrive. I let them sit overnight. The next-morning version of me is much better at telling apart “this critique is right” from “this critique is loud.”
2. Treating Reviewer 3 as a person. Reviewer 3 is a role — the role of the reviewer who didn’t read carefully. Sometimes Reviewer 3 is in slot 1. Once Reviewer 3 was the editor.
3. Counting acceptances. The count I track now is submissions per year, not acceptances. The acceptances will arrive when they arrive.

The hard part isn’t the rejection. The hard part is staying interested in the question after the rejection.