On Mon, 15 Aug 2016 11:31:42 -0400, Keyser Soze
wrote:

On 8/15/16 11:26 AM, wrote:
On Mon, 15 Aug 2016 07:48:54 -0400, "Mr. Luddite"
wrote:


Heh. Yet, you are so quick to criticize Greg with regard to his
qualifications to analyze and interpret statistical data.

Maybe someday it will occur to you that the number of college degrees
one holds is *not* the most significant achievement in life.


I think the difference is that traditionally a statistician was
trained in how to take small snippets of data and extrapolate it into
broad statements about how things are.
I will agree that is a science, bordering on an art.
In the days of "big data" we have the opposite situation. we have
masses of detail data on everything and the trick is distilling it
down into a useable form. In that regard, IBM was way ahead of the
curve. We had guys reporting every hour of their day in 6 minute
increments and every part they used along with computerized dispatch
with every detail of every call and a parts distribution system that
recorded every part from the time it was put into the system until it
was sent to the field. We had masses of data and the trick was making
sense of it. That is similar to what is happening today in other
things.



I'm sure that statistical "training" has adapted since you and I were
sprouts.

If they have actually adapted, a lot of their previous skills are
obsolete.