3 ms·
"X is good for you" is not typically a universal statement, but a probabilistic one. It doesn't have to be good for every single person in every single circumst
by mikeash 11y ago
"X is good for you" is not typically a universal statement, but a probabilistic one. It doesn't have to be good for every single person in every single circumstance in order to be correct. It merely needs to be better for most people.
For example, not being overweight is good for you. The existence of people who survive famine because they started out obese doesn't disprove that.
- IIlllIllIIIIlII 11y ago"Breakfast is good for you" implies that it has some benefit. In some cases, however, the cost of eating breakfast is greater than its benefit. For instance, if you have to get to work at 5am, are you better off waking up earlier to eat breakfast? In this case, the cost of lost sleep may be greater than the benefit of breakfast. The NYT article points out that many of the studies that claim breakfast is good may be flawed. If that's true, then we are likely overestimating the benefit of breakfast. And that means that in many cases where we thought perceived_benefit(breakfast) > cost(breakfast) it could actually be that real_benefit(breakfast) < cost(breakfast)
- mikeash 11y agoYes, of course. I don't disagree with that at all. I'm just saying that "breakfast is good for you" is not disproven merely by showing that it's not always good for you.
- IIlllIllIIIIlII 11y agomy original thoughts were wrong. it makes more sense to use a cost/benefit function (ala economics) rather than a logical argument