I wonder if smoking prevalence among "hackers" is higher than in general population (about 20% aggregate in the US) and over standard deviation over the mean.
Anecdotally and completely unscientifically, my experiences suggest yes, first and foremost as a former smoker who tended to take notice of the distribution of professionals who also took smoke breaks in the half a dozen or so places I worked in three years. Seems like the same sorts of tendencies that make nerds responsive to caffeine probably apply to stimulants in general, especially noting the long-held poetic association between creative, focus-based tasks and tobacco. I also know more than my fair share of ADD types (in a noticeably classical kind of way) who sincerely regard it as a form of viable self-medication.