Edit: And how do you suggest that math isn't odd? 1/5th of the heart attacks were smokers. Roughly 1/5th of the population smokes. That doesn't sound very damning to me.

Your claim was that 20% of the heart attacks were "due to smoking". Not every heart attack that a smoker has is attributed to smoking. You figure out how many heart attacks smokers would have if smoking had no impact, and attribute the excess to the smoking.

Example, using made-up numbers:
1000 corpses. 800 from non-smokers, 200 from smokers.

Heart attack risk in non-smoker: 1/10.
Heart attack risk in non-smoker: 1/4.

Heart attacks (non-smokers): 80. (1/10 of 800)
Heart attacks (smokers, not smoking-related): 20. (1/10 of 200)
Heart attacks (smokers, smoking-related): 30. (1/4 of 200, minus the expected baseline of 20.)

Total heart attacks: 130.
Percentage of heart attacks due to smoking: 30/130: ~23%.

(The RL math is futher complicated by the fact that smokers might avoid heart attacks simply by dying to lung cancer or emphysema first. For example, if heroin caused prostate cancer, you might not see it, simply because heroin users don't often live long enough to die of prostate cancer.)

Not for nothing, but arguing that smoking doesn't shorten your lifespan is more than slightly odd. This isn't exactly a controversial subject among doctors. I don't get where you're going with this.

valguarnera@carrionfields.com