Even if you assume those correlations are 100% accurate, you still have to remember that it's all averages. If I could know that by smoking I'd die five years e

Even if you assume those correlations are 100% accurate, you still have to remember that it's all averages. If I could know that by smoking I'd die five years earlier than otherwise, I'd absolutely keep doing it for the rest of my life. But the chances of cigarettes causing my death exactly five years before I would otherwise have died are very slim, who knows they might give me cancer when I'm 40. Or 90. I know the above is pretty much common sense, but it just always annoys me whenever the topic of "dying an average of X years sooner" comes up.