I suspect this is where it comes down to anecdote and personal preference. Where we both agree I think is that (for programming especially) it's not as simple as X% more hours give you X% more work done; there are other factors at play here..
Now, I could buy a four day work week causing an overall drop in the amount of work done, but comparing them based on the same productivity is meaningless.
Re:Challenging The Math of a Work Week
by yummyfajitas on Wednesday April 25, 2012 @11:40AM
(#3889546)
I'm not comparing based on the same productivity. I'm assuming that boilerplate (done while tired) takes 33% more time.
Most projects involve both boilerplate and cleverness. Mine certainly do.
The cleverness I do while not tired. When I'm tired, and less productive, I do the boilerplate.
Suppose I work 60 hours, with 40 of them occurring before I'm tired. Then I get 40 hours of cleverness and 20 hours of boilerplate (while non-tired, you could probably do the 20 hours of boilerplate in 15 hours).
If you work 40 hours, you are probably doing 25 hours of cleverness and 15 hours of boilerplate (while your 15 hours are equivalent to my 20 hours). That means I'm getting 60% more hours of cleverness than you are.