> make sure that there’s a well-defined distinction between success and failure. Don't fall in the messy middle.

I agree this is the right thing to do, but it isn’t always possible to do for a given experiment. For instance, with the cold-calling example, how do you ensure that it’s either a smashing success or definite failure a-priori? Is there a principled way to choose the threshold number of conversions? Do you just pick an arbitrary number? What if the result is just below your threshold? With some experiments I think the only way to win is not to play. Happy to be corrected if anyone has a solution