I'd love to learn what % of A/B tests get rejected after the test has concluded.

I suspect there's naturally a laundry list of biases that all the work we designed and implemented needs to succeed or boy do we look silly.