Considering the similarities between hacking and writing - highly skilled, solitary, large "page size" - and I think the following excerpt could be enlightening.
From chapter 22, "Professional Writing Expertise" (Ronald T. Kellog), in the excellent Cambridge Handbook Of Expertise and Expert Performance:
...the converging point are that ... all of the idiosyncratic habits of professional writers
(1) focus attention inward by eliminating distractions,
(2) may alter consciousness to facilitate entry in a flow state, and
(3) help regulate the writer's emotional state to keep at the task.
...
Many choose to write at the same time each working day, but individuals differ from morning, afternoon, evening, late night, to early morning preferences. Work sessions of one to two hours correlate with productivity in scientific writing, but the relationship is weak (r=.22) and the variability is large, with some individuals writing four or more hours at a time. Successful poets also typically write for one or two hours, whereas most novelists typically report longer sessions of two to three or even four to six hours. Running or walking help some writers think through problems while away from the writing table. Others use meditation, coffee, cigarettes, alcohol, or other drugs to alter consciousness in the service of writing.
Now, writing isn't the same as hacking, but as a process it seems to have many things in common with hacking. I seem to remember reading somewhere (couldn't find the page) that most experts in general practise in the morning when energy is at its highest, but this doesn't seem to be the case with hacking. I found this rather curious. My guess is that the page size is typically bigger, which requires longer periods of time to enter flow, compared with say, chess, where 30 minutes is often effective. That still doesn't explain why there aren't as many novelists writing at night as there are hackers. So another, less scientifically backed up, idea is that hacking is less of an expertise and more of a "production task", meaning we often do things we already know perfectly well how to do, something that would require less energy. Computer science is an extremely new field, so it wouldn't be too surprising if we would see hackers that are experts at a whole other level 50 years hence. It could also be a cultural thing, since a lot of hackers start as teenagers when they are still in school. What are your thoughts, HN?