>I don't think anyone has claimed that either You said they'd "have to double-run every trial you're doing". Which isn't correct. They have to double run the

>I don't think anyone has claimed that either You said they'd "have to double-run every trial you're doing". Which isn't correct. They have to double run the experiments they screwed up on. That may or may not be faster than taking the time to improve experiment quality.