[Goronmon]

>That's a really scary line of logic though. "We did a thing, and there was no effect. We propose that to obtain the desired effect we should do even more of the thing that had no effect."

Is it scary? I mean, if I go from getting no exercise to spending 30 seconds a day exercising, but see no change in my weight, should I therefore conclude that exercising is pointless? Or should I try exercising more and also maybe be more careful with my eating habits?

I think the scary thing is when people try to take complicated subjects and distill them down to shallow talking points.

[spindritf]

You cannot simply extrapolate from smaller intervention but it does lower the prior for a larger one, of course. Maybe unless you have a specific hypothesis which predicts a non-linear effect.

Also, exercising will actually do very little for your weight because calories have became too cheap and readily available to burn them off. You can eat more by accident than you will use in a fairly vigorous workout.