purplerabbits

How can we get usable statistics from something like this?

You study 1050 people that you selected (they don't say how)
100 of them dies in a fifty year period.
This is said to be twice as many as in the normal population.
1 in 4 of the deaths (only 25 people) are due to drugs and alcohol. So this is what the rest of the article is about.

But what about the other 75? Or rather the other 25 if 50 of them were to be expected? And how many are to be expected in the 'normal population' from drugs and alcohol?

Maybe just as many early deaths on normal people are doe to drugs and alcohol but rockstars are involved in more plane crashes, maybe they are more likely to be depressed and that's why they became musicians. Maybe stadium lights are carcinogenic?

I have no idea, and this article doesn't help in the slightest...