I'm pretty sure the two dishwashers thing has occurred to plenty of people. But dishwashers cost more money than a cabinet and they require a dedicated water li

I'm pretty sure the two dishwashers thing has occurred to plenty of people. But dishwashers cost more money than a cabinet and they require a dedicated water line. This means you can't simply install a second dishwasher in a pre-existing house without first tearing up the walls to add new pipes from your water main to wherever you're going to put the dishwasher, and if you wanted to do this in a brand new house, you'd be asking whoever you're trying to sell it to to pay for an extra dishwasher. And what is with this layperson misunderstanding of placebo effect? Why is this so common? Nobody is trying to subtract it or not induce placebo in real patients. It's the same principle you're applying when evaluating predictive models. You can't simply look at raw accuracy. You need to compare it to some naive predictor to see if it does any better. "Always predict no" is extremely accurate for rare conditions, like "does this patient have ebola" or "is this person a terrorist?" That doesn't make it a good predictive model. Same thing with a treatment. If it does no better than placebo, that isn't to say that placebo is useless. It's to say that we don't need the more complicated, expensive treatment and can simply use placebo. If giving some person a sugar pill has the same effectiveness as giving them a patented synthetic drug with harsh side effect, then just give them the sugar pill. Nobody is trying to avoid placebo. We're trying to avoid unnecessary extra steps.