If you're looking for people who generate a profit from their time, the curve is almost certainly......
The shape of the curve has statistical value and analysis value but most companies are too small
to care about the curve. They are saddled with finding the necessary skill set and also stay
within budget. Some suffer sticker shock.
In my limited lifetime the best model for programmers is book authors.
A programmer must communicate with clarity to the machine and with
management. So as a minimum two context sensitive language constructs
must be mastered.
Skills and details can be learned but the set of necessary detailed skills is moving so fast that
a hiring manager has no clue anymore. When I first put a deck of cards on the in window
for the new 1401 I had a short list of languages to work in. Today I need wc to count them and
all the ways they can be cross linked. N! comes to mind and as the number of coding methods
expands to embrace more languages and vastly more libraries the ability to find a match
becomes astonishingly small.
One might see a list of words unique to one university recolored by the language of the
last programming team. This is seen in patent applications where an examiner has no
encyclopedic knowledge that allows him to see that this is the same as ______.
I have suffered through at least four iterations of RAID technology and tools.
They all had the same underlying "stuff" but the language and tool names
made them incomparable without a magic decoder ring.
The authorship thing is interesting... A look at the NYT best seller list and there
is not enough data to describe the curve. Amazon might have a modern data set
that can describe this but they may not have more than noise.
Good authors of books and programs are simply uncommon.
Editors reviewers and typesetters much more common.